Artem Fokin on Improving with AI and Expert Calls
Fokin sees expert-call libraries and AI as the two disruptive innovations that most changed investment research, not merely incremental process improvements. Walker contrasted independent investors with institutions that could commission 20 calls at roughly $1,000 each and deploy large analyst teams. Fokin says libraries lowered expert-call costs and converted a pure variable-cost model into something semi-fixed and semivariable, while AI compresses summarization work. The emerging edge is “human plus machine,” not either one alone.
An expert call should target the load-bearing assumption in the investment thesis. If the thesis rests on product superiority, interview users, switchers, evaluators who declined, and people who never considered it; if distribution is the question, talk to former salespeople and reconstruct the sales motion. Fokin’s reminder is deceptively important: “Sales just don’t happen.”
The best expert work surfaces disagreement rather than manufacturing consensus. Walker’s medical-device example paired a reported 1% defect rate with 10% for older products, yet surgeons argued that the comparison used decades-old studies and ignored subsequent improvements in technique. Fokin similarly heard reactions ranging from liability-conscious enthusiasm to “I don’t really care” when asking doctors about Sofwave’s almost 10 FDA clearances: “Nothing is certain. Different people have different opinions, and that’s okay.”
Former employees are useful witnesses, but their testimony needs context, corroboration, and relentless follow-ups. Fokin asks laid-off employees whether they would return or recommend the company to family, then checks their claims against product quality, customer enthusiasm, management commentary, and other interviews. Investing rarely reaches “beyond a reasonable doubt”; the achievable standard is a “preponderance of evidence.”
Volume matters when the business is complex, but disciplined curation determines whether volume becomes insight. Fokin conducted five bespoke calls on IWG, still felt he had not figured it out, and then read roughly 70 IWG and WeWork calls; the available library has since approached 100. At Crocs, historical interviews helped distinguish a temporary COVID beneficiary from a company that had built its commercial foundation before COVID while the stock traded around 5.5–6 times 2022 earnings.
Fokin currently gives AI a vote in “digging” and part of “analyzing,” but zero authority in “deciding.” As of August 6, he had found many ways to compress existing work but no meaningful task AI enabled that he could never perform before. A question such as how a company expanded from 10 states can now be answered with historical context and sources in seconds instead of a manual filing search; Walker gave three-proxy incentive-compensation analysis as a similar time saver.
AlphaSense’s Grid and Deep Research workflows turn large document sets into navigable research maps, without replacing primary-source reading. A Grid can place 20 expert calls against as many as 12 recurring questions, exposing consensus, red flags, and the interviews worth reading in full; Deep Research can turn a five-page prompt into a sourced 30–40-page primer that “prepares the mind” before the filings and calls. Fokin attributes “90 or maybe even 95%” of what he knows about new AlphaSense features to repeated sessions with his account manager, while product managers provide another source of use cases and feedback.
1. Expert-call libraries and AI changed the economics of research
Walker’s baseline was the resource gap between institutions and independent investors: McKinsey or Bain Capital could commission 20 calls at roughly $1,000 each and deploy 15 analysts across the literature; a solo practitioner could not. Libraries and AI have narrowed both disadvantages.
Fokin’s correction was categorical: these are not merely “process improvements,” but “disruptive innovation.” Expert-call platforms connected creators and readers, lowered costs, converted a pure variable-cost model into a semi-fixed or semivariable subscription model, and—he believes, without claiming access to the data—expanded both usage and the total addressable market.
Alternative datasets such as credit-card data remain “very, very, very, very expensive” and play little role in Fokin’s process; his simpler alternatives include Google Trends. For him, expert-call libraries and AI are the two largest research innovations of the past decade.
His own expert workflow has two modes: “I either listen or I read.” On IWG, five bespoke calls failed to resolve a complicated global flex-space model, so he methodically read roughly 70 calls spanning IWG and WeWork; with the library now nearer 100, old customer and industry observations still answer questions that recur years later.
2. The thesis determines which expert deserves the hour
Walker’s hard case was a conglomerate such as Berkshire Hathaway—or a roll-up such as Constellation Software—where no obvious expert maps cleanly onto the whole thesis. Fokin’s golf analogy supplied the operating rule: different terrain requires different clubs, so first identify the “key ingredients for success” rather than defaulting to one expert type.
If the claim is that a six-month-old product is materially safer, cheaper, or more efficient, Fokin would concentrate on the product: existing users, customers who switched, evaluators who tried but declined, and prospects who never investigated it. The last two groups can reveal what blocks an apparently “10x better” offering.
If the product is established but growth depends on distribution, the research target changes to former salespeople. Fokin’s Stanford professor reduced business to two problems: “Problem number one, not enough sales. Problem number two, everything else.”
For Sofwave Medical, Fokin asked a former salesperson to role-play a dermatologist visit for roughly 20 minutes: why are you taking time from my patients, and why should I listen? The exercise exposed a sales process that spreadsheets obscure because “it’s very easy to think that sales just happen.”
3. Compliance belongs inside expert selection, not after it
Walker initially described interviewing a company’s dominant customer; Fokin immediately refined that to a former employee of the customer, perhaps someone who left a year earlier. That person might understand the product, procurement process, and relationship without possessing current information that could restrict the investor.
Fokin praised AlphaSense’s conservative compliance process because sourced interviews arrive with a review record. The benefit is not simply access: it is knowing that the expert was screened to avoid an evident legal or compliance problem.
The same discipline applies to every call. The goal is not to extract whatever the expert happens to know, but to obtain thesis-relevant evidence while respecting confidentiality agreements, NDAs, and the boundaries established by the platform.
4. Conflicting customers often reveal more than a clean consensus
Walker’s anonymized implant example looked decisive on paper: an FDA study showed roughly a 1% defect rate for the new device versus 10% for older products, and a defect could require another surgery. He imagined salespeople presenting that as a 10x safety advantage and asking whether a doctor wanted to take on medical-malpractice risk.
Surgeons supplied the missing caveat: the older products’ FDA studies dated to the 1990s, while techniques and processes had improved substantially since then. Their own surgical experience led some to believe the legacy devices were now comparably safe, showing why company messaging, sales execution, and customer belief must be investigated separately.
Fokin found a similarly wide range around Sofwave’s FDA clearances. He believed the number was 9 and rounded it to almost 10 for simplicity. Some doctors valued indication-specific clearance because it reduced perceived liability; others considered it “just a marketing buzz” and were comfortable using devices off-label.
His change with experience was to stop expecting every doctor to return the same answer. “Nothing is certain,” and the point of multiple calls is to map the distribution of opinions, then incorporate that range into the investment judgment.
5. Former employees are witnesses whose credibility must be tested
Walker’s pushback—worth keeping—is that former employees may carry layoff bitterness, while highly successful alumni may remain unrealistically positive even as the business deteriorates. A small sample can contain two satisfied formers and one person for whom a one-star review would be “too high.”
Fokin’s disarming analogy was to former romantic partners: separation does not automatically make their assessment false. More concretely, he asks laid-off employees whether they would return if invited and whether they would recommend the company to a sibling, nephew, or niece; the answers become indirect evidence about culture.
His legal training supplies the standard: investors probably cannot establish facts “beyond a reasonable doubt,” but they can seek a “preponderance of evidence”—a more-likely-than-not case assembled from employees, customers, consultants, management commentary, and observable performance.
Context determines the weight. Because Fokin generally believes unhappy employees and poor cultures do not produce excellent products, a weak product makes a hostile former more credible, while enthusiastic customers and strong products make that testimony less weighty. When someone says the culture was “toxic,” his response is: “Could you give me an example?” Then he asks for another.
6. Screening questions can create value before a call begins
Fokin writes his own three or four screening questions and does not waste one on career history already visible in the biography. He first wants to know what the person actually did day to day, because titles alone can conceal whether the expert touched the thesis-relevant work.
Open-ended questions sometimes generate several revealing sentences before the interview. For a “proverbial needle in a haystack,” he also lists functions—marketing, sales, supply chain, procurement, and others—and asks candidates to rate their knowledge from 1 to 10.
The credible pattern is many ones and twos plus one or two eights or tens. Someone claiming expertise across every function may know little; “marketing, sorry, I have no idea, but supply chain management, nine out of 10” signals both specificity and intellectual honesty.
Interviewers must also respect functional limits: do not ask a salesperson to explain accounting or a marketer why stock-based compensation is high. If Fokin realizes midway that an expert is weaker than expected, he may use the remaining time for broader questions, but the core questions should match what the expert can reasonably know.
7. Historical interviews can separate temporary tailwinds from durable change
Crocs illustrated expert calls as a guided corporate history. Around summer 2022, Fokin recalled the stock trading near 5.5–6 times 2022 earnings, implying either that “the company is going out of business soon” or that it was badly mispriced.
The central question was whether COVID had created a temporary sales spike that would collapse. Former employees—including some who had left before COVID—described internal changes the company had made before the pandemic.
Fokin’s conclusion, explicitly based on his research and therefore fallible, was that Crocs had built a strong foundation before COVID and then used the stay-at-home, casual-wear tailwind to accelerate. The foundation “was probably not disappearing,” making the historical sequence more informative than the headline COVID exposure.
His analogy was a medieval castle tour led by a local history major: a knowledgeable guide connects changes over time that a visitor would not discover alone. Expert calls can provide that chronology even when their subjects lack current operating information.
8. AI accelerates digging and analysis, but judgment remains human
Fokin organizes the research process into Paul Enright’s three stages: “digging, analyzing, deciding.” As of August 6, AI delivered most of its productivity benefit in digging, some in analysis, and “zero in deciding”—a position he expects could change as both the technology and his own practice evolve.
He also distinguishes doing an old task much faster from doing something previously impossible. He has found many examples of the first but “not figured out use cases for the second one yet,” an unusually candid boundary amid broad AI enthusiasm.
In one research example, an old VIC write-up said a company operated in 10 states. Instead of recording a question and later searching filings, Fokin asked AlphaSense for the current count, the history of state expansion, and any relevant revenue breakdown; within seconds, he had the outline and its sources.
Walker offered compensation analysis as the same compression: comparing incentive structures across three proxy statements once required hours of opening, scrolling, and reconciling disclosures. An AI tool can create the first comparison in roughly 10 seconds, leaving the investor to verify and interpret it.
9. Grid and deep research make document overload navigable
AlphaSense’s Grid is Fokin’s favorite feature: it places documents—expert calls, earnings transcripts, filings, or proxies—against as many as 12 recurring questions. He builds templates targeted at products, company strategy, competition, risk, and other research needs; users could also tailor them to SaaS, industrials, consumer, or other styles.
Across 20 expert calls, most with customers, questions such as value proposition, purchase trigger, sales-cycle length, and evaluated alternatives can expose the full opinion range within minutes. The answers also identify the most thoughtful interviews and the yellow or red flags that deserve full, page-by-page reading.
Fokin may still read the other calls to ensure nothing was missed; Grid prioritizes attention rather than replacing source material. His understanding—offered with a non-technologist’s caveat—is that specialized vertical AI also reduces prompt sensitivity by improving a user’s imperfect request behind the scenes.
Walker uses the same structure to challenge earnings excuses: if management blames a “soft consumer,” he compares three peers’ results and commentary before the follow-up call. If peers report strength, management must explain what is company-specific instead of hiding behind the macro narrative.
10. The best AI workflow prepares the mind before primary-source reading
Fokin’s Deep Research prompts can run roughly five pages and return a sourced 30–40-page output covering customers, segmentation, pricing, history, and other recurring questions. He prints it, reads with a pencil, and marks highlights, stars, and margin notes.
The analogy is hearing an hour-long interview with an author before opening a 300–400-page book: the overview gives new facts somewhere to land. Once “my mind is prepared,” Fokin reads the most important expert calls, filings, earnings transcripts, and conference materials with better comprehension and retention.
He had not yet tested AlphaSense’s newly released AI-generated expert calls, so his view remained deliberately open: he was “really curious,” but earnings season had prevented experimentation. That distinction between awareness and actual evidence runs through his broader AI stance.
Feature discovery is itself a research process. Fokin credits “90 or maybe even 95%” of his AlphaSense knowledge to account manager Amar Capellan, including periodic 30-minute reviews every six to eight months; conversations with product managers reveal less-obvious use cases while giving builders customer feedback. His Kasparov-inspired operating belief remains: “Human plus machine is more powerful than machine and definitely more powerful than human.”
Full transcript
You're about to listen to the Yet Another Value Podcast with your host, Andrew Walker. Today's episode is a follow-up to the podcast I did with Artem Fokin, my friend, on perfecting the investment process. This is a webinar that we did on AlphaSense. We talk a lot about using AI and expert calls in the research process, using them for improvement, all that sort of stuff. This was behind a paywall for AlphaSense, but they said after a month, "Hey, we're getting good reviews. Why don't you put it out on the podcast and try and get more listenership and let you know we want this to be out there." So I think you're really going to enjoy it. I hope you do. I'll include a link to the prior perfecting the investment process podcast with Artem in the show notes, which was, to be frank, one of the most popular podcasts I've ever done. We got tons of great feedback on it. So I think you're really going to like that podcast. I think you'll really enjoy this podcast if you like that podcast. So we're going to get to all that in a second, but first a word from our sponsors. Today's episode is sponsored by AlphaSense. Look, over the years, you've heard me talk about it nonstop. AlphaSense, Tegus, they've become core to how I do investment research. They've got a burgeoning set of AI tools. They've got the expert calls, which I absolutely love—the expert-call library. They've got over, as they tell me, over 500 million premium sources from company filings, broker reports, news, trade journals, everything. Plus the expert calls, they put it all in one place. Their AI tools let you search unique data sources in really interesting ways. This October they're hosting their first-ever Alpha Summit 2025 in Brooklyn. I'll be dropping in and out. The event will feature all sorts of leaders from finance—UBS, Wells Fargo, Accenture, Google—who's who are going to be there sharing how AI is reshaping the investment research and decision-making landscape. What's going to make it special is it's not just about the ideas. It's about really talking about how AI can improve workflows and the strategies that top firms are using right now. So I'd love to see you there. If you're going, you can join me there. AlphaSense, Alpha Summit 2025, October 6th through 8th at the Refinery at the Domino. You can sign up at alphasense.com/yavp. That's alpha-sense.com/yavp. All right.
Hello and welcome to the AlphaSense user webinar with me, Andrew Walker, host of Yet Another Value Podcast, and my good friend Artem Fokin. He is the head of Caracal Capital. Artem, how’s it going?
Hi, Andrew. Great seeing you again.
Disclaimer: Nothing on this webinar is investment advice. I don’t think we’re talking about specific stocks, but we might dive into some real-time examples, so people should remember that. I’m sure AlphaSense will do disclaimers out the wazoo.
Look, Artem, let’s just start. I’ll set the stage, and then we can dive into everything. We were talking to the people at AlphaSense. We are both users, subscribers—whatever you want to call it—and I think we both find huge amounts of value from the product. I think we find huge amounts of value in different pieces of the product. AlphaSense and Tegus—they’ve got everything at this point.
We said that, and they said, “Hey, why don’t you guys come on and do a webinar talking about—we just did a podcast talking about, I don’t think it was quite process improvement, but our process as investors.” A lot of the process improvements for me over the past 10 years have been adopting AI into my investing process and adopting expert calls into my investing process. I was already doing a little of both, but both have ramped up materially, in large part thanks to Tegus, AlphaSense, and the tools—the expert-call library, network, and everything.
We told them about all that stuff, and they said, “Hey, why don’t you guys get on and talk about using AI in your process, using expert calls in your process, and how you use AlphaSense in your process?” That’s the overall idea for this webinar: an hour of us talking about all of that. Would you add anything to that, or anything else people should feel like they’re going to get from this hour of the webinar?
I think the theme of both our prior conversation on Yet Another Value Podcast and today’s webinar is the same theme: improving and perfecting our craft as investors. So I stand behind that.
Obviously, when you and I were talking on the public, so to speak, podcast, we were speaking about broader themes as opposed to concrete, specific tools of the trade. We were talking about how to use a hammer to hit a nail, or whatever other tool you may be using. Here, we will probably be talking a little bit more about, “Okay, how do you choose a hammer or any other tool, or how do you use it, and how are you changing your use cases?”
I think this conversation is more likely to be more specific, so I’ll be sharing some of my examples and use cases. You may throw in some of yours. I’m really looking forward to it, but the theme is, I think, the same.
I think you’re exactly correct. I think the way we’re going to structure this is that we’re going to talk about using expert calls, using AI, and then, as we continue, dive deeper and deeper into the how—specifically, how people who are subscribed to AlphaSense and Tegus can use specific AlphaSense and Tegus tools in their process.
Let me just start off. I made a contention there that I’d love for you to push back on, agree with, or disagree with. Over the past 10 years, I think the biggest change and improvement in my process has been this: When I was at McKinsey or Bain Capital, we would do expert calls all the time. But when you switch from a place with approaching unlimited resources to being, let’s just call it, a more individual investor—a small, solo practitioner, whatever it is—you lost that access to, “Hey, I want to learn this company. I’m going to go do 20 expert calls and spend $1,000 per expert call.” You didn’t have 15 analysts to go summarize everything that’s ever been written in a research report.
I think the biggest process improvement has been expert-call libraries, so you can get access to basically unlimited amounts of expert calls for an annual subscription. That’s new over the past 10 years. And then AI can summarize 20 earnings reports, so you don’t need 10 analysts under you to get summaries of big earnings reports. Would you agree with that contention? Would you disagree? How have you thought about those 2 overall process improvements over the past 10 years?
I wouldn’t necessarily call them process improvements per se. I keep using that term, and you keep pushing back on me. I would call them disruptive innovations that came to the investment-research space.
In terms of the tools that I use, those would mean expert-call libraries and bespoke expert calls, because the price dropped dramatically with the advent of Tegus, Stream, Mosaic, and AlphaSense. Again, I will use AlphaSense and Tegus interchangeably because now they’re under the same umbrella, and we don’t need to trace their exact corporate history.
Those lowered the cost, and the library—again, this is technological innovation, a network effect that was built by connecting creators and readers. By the way, you can be both a creator and a reader at the same time. That usually expands the total addressable market and lowers the cost.
Because of that, you convert a pure variable-cost model into a semi-fixed, semivariable model. That would lead to the expansion of the entire TAM and usage of expert calls, in my opinion. I haven’t seen the data, obviously, but that’s my guess, and I think I’m fairly confident that I’m right about that.
So those were 2 disruptive innovations. People may bring in some credit-card data or alternative data that are still very, very, very, very expensive. We don’t use much credit-card data, and our alternative data is pretty simple, such as Google Trends. By the way, AlphaSense, maybe you can figure out how to build something and lower the cost in that vertical as well.
But for me personally, expert-call libraries and AI were the 2 biggest disruptive innovations in the environment that I’ve been implementing into my process.
I agree. Let’s dive into expert calls. The reason I want to start with expert calls is that I think AI is a little buzzier, and everyone should be experienced with AI tools. But I want to start with expert calls because I think they’re a tool that, bluntly wielded, can be useful. The finer you sharpen that sword, the more it becomes a force multiplier in terms of the effectiveness of expert calls.
I mean that both for reading expert-call libraries and transcripts, but especially for conducting expert interviews. I know you and I have done some together before, and I’ve been on calls with other people. When I’m on an expert call with someone who’s really good—someone who’s super prepared—versus someone who might be doing it by the seat of their pants, I’m always impressed by how much more I learn.
I just want to stop there. How do you use expert calls? Then we can start diving into the different ways to use them.
As I joke, most of my time I spend either listening or reading. If you use these 2 modalities, unsurprisingly, these are 2 use cases for expert calls. I either listen, meaning I conduct a bespoke expert call, or I read a certain number of calls that are already in the library.
That number can vary from just a few, if there aren’t that many available, to—I think the record was probably 70 calls across Tegus and AlphaSense that I’ve read. I did probably 5 or 6 of my own as well. I literally read 70 calls.
Since then, the number has only gone up. Now it’s probably up to 100 or so, and I literally read all of them. That was a lot of fun because the business was sufficiently complex. To put things in perspective, it’s IWG, the hybrid-working flex-space provider that operates globally.
The business is complex and had a long history, a lot of evolution, and lots of moving pieces. Full disclaimer: Kraken Capital LLC and all its affiliates own the shares of IWG. This is not investment advice or a recommendation to buy or sell any securities. My first thought was, “I’ll do 5 expert calls and figure this out.” I did 5 bespoke expert calls and did not figure anything out. I was back to square one: “Okay, I guess I need to read these 70 calls.”
Then I methodically read all of them, both on IWG and WeWork, because they’re very close peers. When you study an industry, you want to study several players if they exist. That was incredibly educational and informative for me. Now, when things come up, I’ll think, “Yeah, I know the answer. I’ve read this call.” Someone in the industry or a client will mention something that was discussed 2 or 3 years ago, but the same rationale is applicable today.
I love reading those things, and I also love doing my own calls. In fact, what often excites me when I research a company is logging into AlphaSense, entering the ticker, and choosing to see search results only for expert calls. If it shows 0 calls or 1 or 2 calls, it probably means that nobody has really looked at it. For me, that’s an opportunity to learn more about a company that I perceive as an interesting investment opportunity and deepen my research tremendously compared with what’s probably out there.
Let me start with one question: sourcing experts. I run the Yet Another Value Podcast, and one thing I’ve started doing recently is trying to do an expert call on every company before I do the podcast on that company. It’s an idea-focused podcast, and the episodes are about an hour long. I’ve started trying to do an expert call on every company so that I have more informed questions and more interesting insights.
One thing that’s jumped out at me is that, for some companies, it’s obvious who I want to do the expert call with. If you’re doing a company that has 1 customer representing 80% of its business, I want to talk to somebody who works at that 80% customer and get their insights into the company. If you’re doing a company that has a new drug, I want to talk to a doctor who’s prescribing the drug and ask how they’re viewing it and how they’re seeing it.
For some companies—I’ll use this as a very loose example—if I were going to do an expert call on Berkshire Hathaway, what expert call would I do? It’s a conglomerate that is basically Warren Buffett. A former employee probably wouldn’t really help. Charlie Munger, rest in peace, wouldn’t even talk to me, but I wouldn’t do a former-employee call.
It’s really a culture concept. I say that because Constellation Software is a roll-up of software businesses focused on municipal governments, if I remember its core focus correctly. What am I going to do with Constellation Software if I’m going to talk to an expert? It’s literally 100 different small companies rolled up together. There are several other examples, but I struggle with that. When do you know if an expert call works for a company or works for an investment thesis, versus when it doesn’t, like in the Berkshire Hathaway example?
Okay, great question. Before I answer that, let me clarify something. I think what you probably meant to say was that if a company has a very large customer, you may want to talk to that customer. What you meant to say is that you want to talk to a former employee of that very large customer.
Yes, exactly.
Because from a compliance perspective, talking to a big customer is not prudent or smart. You’re right: the compliance people will jump on that. I want to be very clear that AlphaSense has a fantastic compliance team. They’re very thoughtful, very diligent, and conservative in a good way. They would not source an expert who could present a compliance or legal issue.
In Andrew’s example, the right way to go about it would be to find someone who left that customer a year ago.
Yes, someone who may understand the product, understand the process, and understand the relationship, but who has no information that may restrict the investor.
Yes. AlphaSense’s team is fantastic in that sense. That’s another great thing about using AlphaSense’s expert-call library or asking AlphaSense to organize calls. There’s a record showing that the call has been reviewed by the compliance team, and you know that you’re clear and good to go. That’s another important point. It’s not what you asked, but it was very pertinent to the point you made.
Going back to the heart of your question—sourcing expert calls—remember our conversation about golf and different clubs yesterday?
Yes.
For different situations, you use different tools. Similarly, in golf, for different terrain, you use different clubs. Otherwise, you’re not going to do well at golf. The same principle applies here.
Usually, you figure out the key ingredients for the success of your investment thesis. For example, if your idea is based on product superiority—that the company launched a product 6 months ago and you believe it’s massively superior to everything else out there—maybe it’s safer, cheaper, or more efficient. People in Silicon Valley like to say, “This is 10x better” or “10x cheaper.” That’s a figurative way to describe it; it may not necessarily be 10x better, but it should be substantially better.
In that case, I would want to focus all my research efforts on that product and understand the customer perspective. I would talk to people who are already using that superior product, people who switched to it, people who tried it but did not switch, and people who didn’t even bother to learn about the product. I’d want to understand what’s stopping them. That’s one example.
Another example is a thesis based on the idea that the product is fantastic, and I’m comfortable with that. In business, there are 2 problems, as my Stanford Business School professor joked. Do you know those problems?
Problem number 1: not enough sales. Problem number 2: everything else.
Okay, so you have the product. Can the company sell a lot of that product or service? Do they have a well-established sales motion and an incredibly well-running sales machine, or not? In that case, I would want to talk to former salespeople.
What is the sales process like? There’s another company that Kraken Capital LLC and all its affiliates own called Sofwave Medical, which is listed in Israel. Again, this is not investment advice. In my opinion, based on my research, they have a superior product. You can ask me later how I reached that conclusion and why I think it’s superior, but I also wanted to understand the sales process.
AlphaSense helped me source a former salesperson, and I spoke with him for an hour. I literally asked him, “Could you role-play a conversation with me? Imagine I’m a doctor, a dermatologist. You came to my office, we shook hands, and I asked you why you’re here and why you’re taking time away from my schedule and my patients. Let’s go through that.”
We did a role-play for probably 20 minutes over the course of the hour, and then I asked various other questions. That’s incredibly helpful for understanding how these sales happen, because when you’re in Excel, Notion, Microsoft Word, or whatever we use, it’s very easy to think that sales just happen. But sales don’t just happen. Someone needs to sell the product. Everything in my office here has somehow been sold to me.
I believe understanding the sales process is one of the most incredible things. If I look back on my life—I’m in my early 40s—I wish I had spent 6 months in sales, ideally in a business-to-business role, just to understand the process, its psychological challenges, and how it works. I think it would have made me a better investor.
I’m in my early 40s now, so I’m not going to go back and get that job. The best I can do is either talk to people who are in sales and learn from them or read expert-call libraries about the sales process and how it works.
I think you’ll know the company I’m talking about. I’m going to keep it anonymized intentionally, but one really interesting example I’ve seen recently is a company with an FDA-approved study showing that its device defect rate is 1%. All of its peers use last-generation products, and their defect rate is 10%.
When you see that, you say, “Oh my God, this product is literally 10x better.” You chose 10x. These defects are pretty bad—it’s an implantable device. If you have a defect, another surgery is required at a minimum, right?
You read that and say, “Oh my God, this is a 10x-better product.” It’s so interesting when you do expert calls on this company and talk to salespeople about how they’re selling it to people using that message: “Hey, this product is 10x better.”
Do you really want to take the medical malpractice risk of putting a different product in when this product is 10% better? Compare that with when you talk to doctors and say, “Hey, I’m reading an FDA study that says this product has a 1% defect rate versus other last-gen products at 10%. Why are you still using a 10% last-gen product?” And they say, “Hey, all of the FDA studies on the last-gen products were done in the ’90s. Guess what? There have been huge process improvements and huge improvements in surgical techniques since then. We think that the last-gen products are just as safe as the current-gen products, and we’ve got our own evidence from our surgeries to back it up.”
It’s really interesting. You hear the company spin, and then you can use the expert calls to dive in and hear how the sales force is using it, and maybe whether the customer—in this case, the surgeons—is buying it or not buying it. I have more questions on expert calls, but I think that’s a really interesting anonymized example I tried to give. I’ll pause there and let you comment on any piece, or comment on that again, on a company that we own and that has FDA clearances.
We still own the shares. Nothing changed in the last 5 minutes. They have almost 10—I believe the number is 9—so let’s round it to 10 for the sake of simplicity, as the clearances for various indications. As far as I understand, based on my research and conversations with AlphaSense, almost nobody else has as many.
So then I ask doctors, “Sofwave has this number of clearances, and that device doesn’t, or has only 1 or 2. Do you, doctor, care?” Interestingly, I receive a fairly broad range of responses. Some people care because they say, “My liability if something goes wrong is a lot lower because I was using an FDA-cleared device for that indication. It gives me more comfort. I like that.”
On the other extreme—and I’ll skip all the shades between those 2 extremes—someone may say, “I don’t really care. I’m really good at what I do. I can use it off-label. I’m a super-confident doctor. I don’t care. For me, it’s just marketing buzz.” That’s a range of opinions.
And that’s something I’ve gotten more comfortable with over the years, as I’ve done many, many expert calls: Sometimes there will be no one answer because different customers may have different opinions, and that’s okay. You go in expecting that when you ask a question, every single doctor in our example will give you the same answer.
Yes. Yes.
Great, I got it. It’s certain. Nothing is certain. Different people have different opinions, and that’s okay. But it’s incredibly helpful to understand the broad range of opinions and incorporate them into your thinking about the investment thesis, whatever that thesis is.
Let me go back. You mentioned, when you were talking about Sofwave, talking to a former salesperson. Former employees are actually the discussions I have the hardest time with because, by definition, they’ve left the company.
I think everybody left the company; otherwise, they’re not former employees.
By definition, that’s why I said they’ve left the company. Most of the former employees I talk to have left through a layoff, especially now, on the heels of 2022, with all the growth stocks going through a round of cost-cutting and fat-trimming. A lot of the former employees left through layoffs, and I find that sometimes it is hard to cut through the bitterness of having left through a layoff versus how the customers actually treat the company.
Or sometimes the person loved being at the place, was wildly successful, left, and has nothing but great things to say about the company. Then you look and the company is kind of going up in flames, and you’re like, “Oh.” It’s hard to reconcile those things. I want to ask: How do you separate the bias of a former employee, whether it’s positive or negative, from what you’re learning in the call? I just find they can be such double-edged swords when you talk to them.
It’s difficult. Let me ask you this before I answer the question. As far as I know, you’re married. I know your wife’s name. I think you mentioned it publicly. I’ll say her name is Alicia. So, okay, fantastic. Have you had girlfriends before you married?
Art, are you trying to get me in trouble here, bro?
No, no, I’m not. I hope—
I might have had 1 or 2.
Okay. I’m hoping that if someone calls them and asks, “Is Andrew a good human being?” they will probably say that you’re a good human being, even though you’re not together anymore. So I think former employees are a little bit the same.
And, by the way, former employees leave for a variety of reasons. Not all of them got laid off. Some of them got a better career option.
Yes.
They accepted the offer and moved on. There aren’t really many bitter feelings in that case. There will be some thoughts. But where I’m leading with my question about your dating history—Alicia, I’m really sorry. I hope I will not get Andrew into trouble—is that I think how the company would treat a former employee, if that employee was dismissed, whether fired or laid off, also tells you a lot about the company culture.
One of the best indicators—indirect evidence—about culture that I’ve gotten from former employees is when a person will honestly tell me, “There was a round of layoffs. Unfortunately, I was let go. I think it’s a great company. It’s unfortunate.” Then you ask them, “If they called you tomorrow and said, ‘Hey, business is doing better. We need more people. We love you. We’re sorry that you had to leave. Would you like to come back?’” Many people say, “Oh, yeah.” Some people say no. Some people say yes. That’s a telling sign.
Similarly, would you recommend your brother or sister, or a nephew or niece, to work there? It tells you something about the culture. Now, it’s only one piece of evidence. Again, you and I spoke about this yesterday: We are in the business of making judgments and decisions. Similarly, in this case, when we talk to any expert, we need to make a judgment about their credibility. It’s not always easy. Often, it’s difficult.
You know what I was doing before I went to business school, right? In my prior professional life—
Yes.
Okay. What I was doing before that—
You were a tax lawyer.
Yeah, I was a lawyer, right? In law, there is this concept of a preponderance of the evidence. I think some of those frameworks from my legal days are still mental frameworks that I use in my investing.
A preponderance of the evidence would be one of those. I cannot establish with certainty that what 10 experts—customers, former employees, industry consultants, whoever—told me is true. I don’t know. They may be lying; it’s possible. They may be telling me the truth, but they’re simply wrong. It could happen. All of us are human beings.
But what you’re trying to do is build a case where there is a very high chance, supported by that preponderance of the evidence, that this would be a good investment to make based on your thesis.
Can you ever get beyond a reasonable doubt, which is the standard in criminal proceedings? Probably not in investing, unfortunately. But could you get at least to a preponderance of the evidence—the “more likely than not” standard? I think so. That’s what we’re trying to do.
It would be awesome if we could interview 5,000 former employees at every company, really build a database, and say, “Oh, yeah, there were 20 who were upset, but 4,980 weren’t. This is a great company.” In terms of extremes, I’ve done calls before, and I’m thinking of a specific company with 3 former employees. This was a smaller company, so 3 former employees was a large percentage of their former employees.
Two of them were happy and had good things to say, and one of them was so negative that a 1-star review would have been too generous. It was as little as they could give. This was the worst place in the world. All of these people had left through layoffs, and I don’t know—the person I was interviewing had been there for 9 months. They might have been surprised by the layoff. It was a pretty miserable experience.
But I wonder how you weigh the extremes. Whether it’s my example, where there’s one person who is unbelievably bitter, or sometimes I’ll do calls where 3 people are ambivalent—the company is fine, they wouldn’t go back, but they have no bad words to say—and 1 person is over the moon. In investing, extremes are what make much of an investing career. Do you weigh extremes more or less? How do you think about that?
I need context. I cannot apply this in general without reading those transcripts and making a specific judgment call. That’s number 1. Number 2, you also get other data points.
For example, this is my frame of reference, which may be incorrect, but I believe that generally, companies with crappy cultures and unhappy employees do not make great products. Usually, it means that if I’m seeing that the company has a really crappy product, and 1 out of 3 employees in your example was unhappy and spoke very negatively—a 1-star review would be too generous—
I'm more likely to believe the expert testimony—let's use that word. If the company is shipping great products and has happy customers who are enthusiastic about the product or service, I'm probably less likely to put more weight on the unhappy former employee.
And, by the way, all of us have this type of experience. A few years ago, I invested in a health care services company—that's not a current position. The health care services company, broadly defined, was more focused on dealing with insurance companies, let's say. I spoke with a few former employees, and one of them was very, very negative. But when you look at his specific lines and what he said, and match it with management commentary and what other people are saying, you get to the conclusion that he's probably off base, too emotional, and too unhappy, and he's not really supporting it with facts.
Then there is another thing that is important here: the art of asking follow-up questions. If someone says, “The company culture is horrible. It's horrendous. It's toxic,” you're very interested to hear more: “I'm so glad you said that. Could you give me an example?” If you get 1 example, you ask for another example; you keep digging. Think about it: you put that person on the witness stand while respecting their confidentiality agreements and NDAs and everything else. Of course, you try to get to what's behind the statements. If they give you examples that seem totally toxic to you, maybe there's a problem. If they cannot come up with any tangible examples and it's all big words, they probably aren't credible. That's great on expert calls.
Again, the reason I wanted to start there is because, at this point, you and I have talked about it. I do an expert call a week. I probably read a transcript a week. I find them to be hugely helpful tools, and the more you apply them to investing—even if you're following a company and calling a customer, a different customer, once a month or once a quarter, whatever it is, and asking them, "Hey, you're buying this product. What do you think?"—obviously within all compliance rules, of course—it's really interesting to continue to build that mosaic. If people aren't doing it, I think they're getting left behind. It's one of the few areas where you can build mosaic information that's going to be pretty unique to you. Any last thoughts on expert calls, or can I start switching over to AI?
I have a few points on expert calls. First, screening questions alone can be a fantastic source of value.
Yes. So what I usually do is design my own expert calls and ask the team, “Could you ask these questions on the call?” I never ask anything about career history because that will be available in the bio. You don't want to waste a screening question on that because you cannot ask 20 of them; you ask 3 or 4. That's number 1.
Number 2, I always try to understand what that person was doing day-to-day. Sometimes it's pretty obvious; sometimes it's less obvious. I try to ask questions of 2 types. They can be open-ended, and I'm hoping that the expert will write 2, 3, or 4 sentences. Sometimes you learn a lot from those sentences alone.
Some questions apply particularly when you're looking for a proverbial needle in a haystack. You ask them to rate their knowledge of several areas of the business—marketing, sales, supply chain management, procurement, and so on—on a scale of 1 to 10. Ideally, you want to see a lot of 1s and 2s and 1 or 2 8s or 10s, because that would mean the expert is honest and not exaggerating their expertise. If they say, “I can talk about sales, technology, product management, and marketing,” they probably know nothing, unless they're a CEO. But if they say, “Marketing? Sorry, I have no idea. Supply chain management? 9 out of 10. I'm very knowledgeable. That was my title; that's what I was doing,” that's a very different game.
When I read a lot of expert calls, as you and I have discussed, sometimes it's very clear that the people conducting them ask questions that the expert won't be able to answer. For example, don't ask a salesperson about accounting, or ask a marketing person why stock-based compensation is so high. They wouldn't know unless they're an accounting junkie who studies accounting textbooks for fun on Saturdays and Sundays. You need to calibrate that.
Sometimes, I get it, you may have a call for an hour. You're 40 minutes in, you've asked all your questions, and you realize the expert maybe isn't as good as you hoped, which happens. Then you say, “Okay, let me ask everything else.” It could happen. But sometimes I see it and think, “Oh my God, why are they asking this? They should have followed up on this.” I'm sure I've made my own mistakes. I'm not saying I'm perfect whatsoever. That's continuous improvement, but that's another common mistake.
One use case I wanted to highlight from a prior investment at Caracal Capital, which we don't own shares of anymore, is a company called Crocs—the very interesting-looking shoes, as I joke.
Let me set the stage. As far as I remember, in summer 2022, Crocs was trading at probably 5.5 or maybe 6 times earnings for 2022, and it was already, let's say, June. It doesn't take a genius to figure out that either the company is going out of business soon, in the next few years, or it's horribly mispriced.
Crocs had a tremendous uplift in sales during COVID, and the key question was: Is it just another COVID beneficiary and is it all going to collapse? By collapse, I don't mean the stock price; it was already down horribly. I'm talking about fundamental performance and whether it was going to collapse. How do you figure this out? It's very difficult.
I spoke with several former employees, and some of them had left even before COVID. They shared with me all those internal changes that the company had made before COVID, and then COVID probably helped with the whole stay-at-home and very casual-wearing trend. Sure. But the company built such a fantastic foundation—based on my research; I can be mistaken—that they were able to take advantage of those COVID tailwinds, and that foundation probably isn't disappearing. That was a very valuable learning experience.
Again, that was almost a history lesson: “Look, this is what happened in 2017; this is what happened in 2018. This was the change. We got the new management in place, we changed this, we changed that,” and so on. That was an incredible history tutorial. It's almost like traveling to France to see some castles from medieval times and getting a local guide with a major in history. Guess what? They will tell you a lot of valuable things that you would not discover otherwise. That's the analogy.
That was great. I actually have follow-up questions, but I want to be cognizant of time on the webinar and everything. I want to move on to talking about AI, and I'd love to just start broadly. We can dive into AlphaSense-specific tools, and I do want to talk about AlphaSense-specific tools since this is an AlphaSense webinar, but just broadly, how have you been incorporating AI into your process? How are you using it? What tools are you using?
Okay, so I think this is where we stopped yesterday. The way I think about the usage of AI in my investing process is across the lines of the framework that I've heard from Paul Enright on one of his podcasts. I believe it was a podcast with Patans on Invest Like the Best. If I'm wrong, then it was on Capital Allocators.
He breaks it into digging, analyzing, and deciding. It's very simple. There are 3 stages. It's not a flowchart diagram with 100 boxes; it's pretty simple, but I found it's effective in general and especially when I think about AI.
I break it down by where AI is delivering the most productivity boost for me. It's mostly in digging, a little bit in analyzing, and so far 0 in deciding. This is my view as of August 6, I believe—that's today. It could change in the future as technology evolves and as I evolve.
Simplistically, we can break the usage of AI into 2 questions: Can you do something with AI that you used to do, but it will save you a lot of time? Or can you do something with AI that you were never able to do? I've figured out many use cases for the first one. I have not figured out use cases for the second one yet. I hope I will.
On digging, AI provides a productivity boost in many ways. I'll walk you through some use cases. As I mentioned yesterday, I feel that AI and expert calls are a match made in heaven. Now I can use AI as a core research copilot.
I'll give you an example. I met a company at the B. Riley conference in May. It was an interesting group meeting, and I came away thinking that maybe it was an interesting company to research. I got back home a few days later and pulled up some VIC write-ups just to understand what the thesis had been.
I started reading the write-up, and it said this company operates in 10 states. There are 50 states in the nation, so they still had 40 states left—at least based on information from roughly 2 years ago. But had that changed? It had been 2 years. Maybe now they're in 30 or 50—or 60. Okay, not 60. 50.
In the old days, what I would need to do is open my Notion file, type in the question list that I have in the template, and type, “How many states is this company currently operating in?” Then I would need to open the 10-K and look for it.
Etc., etc. Now what I do is have my PDF file with the VIC write-up, or VIC itself, on one screen, and AlphaSense on another. I pull up the AI assistant and ask, “How many states is the company currently operating in? Also, could you please provide the history of how the number of states changed over the years and any other relevant information, such as revenue breakdown by state or group of states, or whatever?”
I don’t know—5 seconds, 10 seconds, 8 seconds—and I get an answer. I quickly read it and know, “Okay, since then, they’ve expanded to this number of states. They’ve got core states, or legacy states, where most of the revenue comes from, and they’ve got new states that they’re still ramping up.” Fantastic. I don’t need to write this question down anymore and then remember to check the 10-K. So that’s what I need from a research copilot.
I love that. I’m completely with you, just in terms of the use cases for summarizing. I know I want to go to the past 3 proxies and see how incentive compensation has changed over time. That used to be a multihour process, right? You have to open 3 proxies, scroll through them, really compare them, and read. These companies don’t exactly make it easy to see how much they’re paying each person and why they’re paying them.
With any type of AI tool, it’s a second—a 10-second process. “Hey, how has incentive compensation evolved over the past 3 years?” Let me ask specifically: that was just generic AI. I think there are really interesting things about merging AI and expert calls, but I think we can hit on them as we talk. What about AlphaSense specifically? Especially over the past 6 to 9 months, they’ve rolled out several tools, they’re going to roll out several more, and they’re getting much better. What AlphaSense-specific tools are you using?
Okay. So you’re right. AlphaSense is shipping a bunch of new features, right? That’s fantastic. Features are getting shipped and shipped and shipped, and there are more and more and more. The one they obviously have is Chat. Chat can work as a regular chat, or you can turn on Deep Search mode in Chat if you want a bigger output and you want it to think longer. That’s fairly similar to the use case for ChatGPT, Perplexity, or whatever general AI—or what I call horizontal AI—you use.
My favorite personal feature of AlphaSense is what they call Grid. You can make Grid almost whatever you want; it’s like a blank slate. I’ll give you a few examples that I use, and I’ll give you some examples that I think I would like to use but haven’t done much of yet because of my investment style, as opposed to limitations of AlphaSense, because that tool is very powerful.
Imagine Andrew calls me and says, “Artem, I have this great stock idea. This is the thesis,” etc. Okay, I want to understand the customer perspective. I go to AlphaSense, and there are 20 expert calls, most of them with customers. I have no clue which one was a good call and which one was an average call, and all of them are pretty long. Twenty calls is a good amount of time to read, and usually, when I read, I also take a lot of notes, which means my reading time actually goes up. Figuring out where to focus—and this goes back to the digging stage—is very, very important.
Now, what I have is prebuilt—and this is important. AlphaSense gives you some templates; you can use them, but I made those templates myself to fit my investment style and focus. I have different templates for Grid. I’ll have templates targeted more toward the product, and another one targeted more toward company strategy, competition, risk, and so on. You can build as many as you want. If you’re a software-as-a-service investor, you can build one for SaaS; another one for industrials or consumer—whatever you like. There’s a lot of flexibility there.
Grid looks exactly like a chessboard. There are horizontal lines and vertical lines. The horizontal lines are documents. By the way, documents can be anything: expert calls—you know, that’s what I’m using as my example—earnings calls, 10-Ks and Qs, or proxies, as in your example about incentive compensation. Whatever you like. The columns are questions, and I think you can put up to 12 questions. You ask those questions and see what you get in the grid.
That’s another cool thing about AlphaSense. I’m not a technologist, so if what I’ve described about AlphaSense is actually incorrect, I apologize in advance. But my understanding is that AlphaSense, because it’s a specialized vertical AI tool, makes prompting by the user less important and less critical for getting a good output. I think that, at the backend, they transform your lousy prompt into something a lot more thoughtful. You can still try to be thoughtful, and I try to do that, but that’s a big advantage of using a specialized AI tool versus a generic or general AI tool such as ChatGPT, Perplexity, or whatever else.
I might ask, “What’s the customer value proposition? What made you decide to buy this product? How long will the sales cycle be? What other alternatives have you considered?” Then you click; it takes a little bit of time, and you get this grid with 20 expert calls horizontally and your 12 questions vertically. You can read very quickly through 20 answers, if they were given, because some of them may not have covered a particular question. You just move along the vertical line.
Number one, you get 20 points of view in just a few minutes. Remember, you and I spoke about doctors who may give different opinions about the same issue. Now you’ve got the entire range of opinions within 5 minutes, 10 minutes, or whatever it takes for you to read them. More importantly, you can repeat this for all 12 questions. You can also figure out which expert, or several of them, give the most thoughtful answers.
Yes.
Alternatively, I can figure out which expert call gives me yellow flags or red flags. If I see, let’s say, out of 20 calls, 3 with red flags or yellow flags, I’ll go there and read them myself from the first page to the last page, including the disclaimer and page numbers. Then I may decide that, given these yellow flags, I’m not comfortable making this investment, so I’ll kill my research idea.
Alternatively, I’ll figure out 3, 4, or 5 calls that are the best and go there. Then I may decide whether the other 15 are not important, or whether I’m still okay. Knowing me, I’ll still read them, but it would be just to make sure that I didn’t miss something, because my thesis has already been confirmed. That’s a massive boost in time and efficiency driven by AI.
The theme here for me is that human plus machine is more powerful than machine, and definitely more powerful than human. I think I got this idea from Garry Kasparov, the 13th world chess champion. He coined it many years ago, and the premise is that human plus machine is more powerful than machine and definitely more powerful than human. That’s the idea. It may not always be the case—who knows? We’ll do another webinar in 10 years where my avatar will be talking to Andrew’s avatar, all powered by AI tools. Maybe I’m wrong on that premise, but that’s how I approach it.
The other interesting thing I found, very similar to what you said, is that I like to ask a question, especially when I’m new to a company, in Generative Grid. Then I just look through it and see that it labels which call the quotes are coming from. Often, one call will have the most quotes popping up, and I’ll think, “Oh, that call is the one most likely to answer the question that I’m asking.”
So I don’t have to read through 6 of the 20 transcripts until I find, “Oh, here’s the one that really talks about Salesforce.” It’s just there. That might sound obvious, but it often isn’t obvious when you’re just looking at the overall picture. I’ve found that’s been a great way to really speed up what I want to focus on.
One thing—let me ask ahead: can I talk about Deep Research a little bit further? I mentioned that in passing, but I think it’s very much worth paying attention to. This is how I mostly use the Deep Research feature.
Imagine I get an idea from somewhere. It may be a screen, it may be Andrew’s podcast, Yet Another Value Podcast, or it can be a peer in my network who shared an interesting thesis with me. Whatever the source is, my old process would be to go read the 10-K, or read several earnings calls and conference call presentations, such as Morgan Stanley TMT or J.P. Morgan Healthcare, or whatever the case may be.
I would slowly build a mosaic in my head, and at that point, even after reading those 5 calls, it would be far from perfect because I’m trying to put different pieces into the right places. Who is the customer base? What’s the segmentation? What’s the pricing? It’s slow, plus the corporate history.
The second input for what I’m going to say as a conclusion is this: think about it as reading books. If you read a book first and then reread it 3 months later, you’ll probably get a much deeper understanding of more intricate and complex concepts the second time than you did the first time. I believe the same applies to reading investment materials. If my mind isn’t prepared to absorb a lot of information and immediately put it into the right places, the right places here are the key.
I will probably create a little bit of a mess in my head, and it will take me more time to sort it out. But how can I apply that concept of reading a book first and then reading it 3 months later? A better analogy would be this: You get a great book recommendation. You go, “I’ll listen to a podcast with the author for an hour, get the key concepts, and say, ‘Okay, my mind is prepared.’” Then I’ll go and read the book, and my level of comprehension and retention will be a lot higher.
The same applies to investing. That’s what I’m trying to replicate. I have a few prompts that I will use for the Deep Research feature. It’s a pretty long prompt, probably 5 pages, and I will ask AlphaSense’s Deep Research to return a pretty detailed output that will cover a lot of my questions. Then I will sit down. I usually print it out because that’s how I like reading.
I sit down in my reading chair, get a pencil, and go through it—highlighting, writing something in the margins, putting stars, scribbling remarks, et cetera. The usual output is 30–40 pages, would be my guess. It will have great sources, with references to expert calls, earnings calls, or conference calls. Then, as Andrew said, you can read the sources that are most frequently cited.
After those 30 or 40 pages, my mind is prepared. It’s equivalent to listening to a podcast with an author for an hour before reading the entire 300- to 400-page book. Then, with a prepared mind, I will start reading all those primary documents, or I will start with the most important ones. That’s how I use Deep Research.
I also want to mention the new AI feature that I think just got shipped. I haven’t even used it yet, but I’m very curious to try it. I think it’s AI-generated expert calls with experts, where AI is asking questions. I haven’t had a chance to test those. They were released very recently, and with earnings season happening right now, I didn’t get a chance to try them, but I’m really curious.
Well, that actually brings me—I want to be cognizant of time because it’s an AlphaSense webinar, not my podcast where we can ramble for an hour—to what I thought was going to be my last question. Maybe we’ll make this the second-to-last question.
In the same way Bloomberg was, if anyone’s had a Bloomberg terminal, they’re always releasing features, and everyone knows they only use about 5% of the Bloomberg features. AlphaSense is releasing features really quickly. I didn’t even know about this expert AI transcription feature. I use AlphaSense pretty extensively, but I may not have seen it. Maybe I wasn’t on the beta. I don’t know.
What have you found to be the best way to keep up with new tools? With AI tools, many of them are much different from tools we’ve ever used before. You get this new tool and you might think, “Oh, that’s free. I don’t know.” It might be the most powerful tool ever released. How are you keeping up with AI broadly, if you want, but preferably with AlphaSense in particular and all the new features that they continuously roll out?
Ninety or maybe even 95% of what I know about how to use AlphaSense, especially the new features and new use cases, I’ve learned from Amar Capellan, the account manager responsible for Caro-Kann Capital’s account. He’s incredibly helpful. Amar, I’m shouting out to you: Thank you so much for teaching me everything, and I apologize for asking a bunch of dumb questions. Thank you for your patience.
I’ve learned from my account manager. I will ask, “Oh, this feature was released. Would you show me how to use it?” Also, because I’ve been working with Amar for a number of years, I think he has gotten to know my investment style and what I do and don’t want to do reasonably well. He will point out, “I think you’ll like this use case for this feature.” I’ll say, “Oh, that’s great. I do. That’s fantastic.”
That’s number 1. Number 2, sometimes I will ask Amar to spend 30 minutes with me on Zoom and do a regular catch-up call—maybe every 6 months, maybe every 8 months. I’ll ask, “Hey, do you have any new things that you changed, improved, or added? Could you show me?” That’s another way for me to get up to speed and know what’s changing.
For example, Amar showed me how to use Generative Grid and the different use cases for expert calls or earnings calls. Another cool feature of Generative Grid is that, for example, let’s say you cover 20 consumer goods companies and you’re trying to figure out what they said about macroeconomic consumer confidence. You can pull that into Generative Grid. I don’t do it as much because I’m not necessarily an industry specialist, but it’s another great use case.
I have found it really, really useful. Every company, when they pull guidance, miss guidance, or reduce guidance, says, “The consumer is soft.”
Most people do follow-up calls with management teams—not all, but most—12 hours, maybe 24 hours, maybe even 12 minutes after the earnings call ends. I’ve found it so useful to be able to ask, “Hey, this company said that the consumer was soft during the quarter. Here are X, Y, and Z, their 3 peers. What did they say?”
Then you get on the call with management and say, “Hey, you guys said that the consumer was soft. X, Y, and Z over there all had better results than yours. They said the consumer was strong. What is it about your company that means the consumer is soft? Are you guys just using an excuse? Is there something about your specific consumer?” I’ve found it’s just so good.
That’s something you can do on your own, but if you want to do it fast, you can’t do it fast. It takes a lot of time, and it’s a great summary. Then, if you want to dive into it, you can say, “Okay, X, Y, and Z—this is what the Generative Grid said. Let me dive in and see specifically what they said.”
Last thing—we’re running out of time. You talk to your account manager a lot, and you use AlphaSense a lot. I know that I’m not using all their products properly. What is one thing that you’re getting a lot of value out of AlphaSense that you think maybe the average user underutilizes or doesn’t realize is out there? What’s one AlphaSense tool that you would recommend for that?
I don’t necessarily know what a typical user may or may not be doing, so my best guess would be this. Over the years, I’ve spoken with a number of product managers who are responsible for building specific features. I’m talking about broad features, like the Notebook product manager or other broad feature categories—not small ones.
I’ve spoken with a number of them over the years, and I think it’s also a great way to learn about potential use cases, including very powerful use cases that may not be obvious to a user right away. The people who were running the entire process—from ideating to building a prototype, testing, QA, quality assurance, and shipping to users—are probably the most knowledgeable people about that feature.
That’s pretty cool. I’m also hoping that it was a way I was useful to those people who generously shared their time, at least somewhat, at least a little bit. It’s also a way to get the customer’s voice. Sometimes you may say, “Do you think you can add this thing, this wrinkle? It would be really, really cool.”
Over the years, probably some of those thoughts from customers like me got implemented. Some thoughts from me may be too specific to my style and not relevant for many others, and they probably would be ignored, which is totally fine. But some, I hope, would get into the product because a few users like me would speak up about those things.
That’s my best answer. I’m not sure whether many other users of the AlphaSense platform necessarily do that well.
No, I think that’s great. I’m certainly not reaching out to the product managers specifically.
Okay. Anyway, we’ve run for a little over an hour, so I think we’re really breaching the limits of what the AlphaSense webinar platform can handle. Artem Fokin from Caro-Kann Capital, both of us are power users and happy users. Thanks for hopping on this podcast and discussing all things AI, expert calls, AlphaSense—everything. It was really fun, and we’ll have to chat soon.
Okay, talk. Bye, Andrew.
Bye, everybody.
A quick disclaimer. Nothing on this podcast should be considered investment advice. Guests or the hosts may have positions in any of the stocks mentioned during this podcast. Please do your own work and consult a financial adviser. Thanks.