Sohn Conference Foundation · · 9 min
Rahul Kishore pitches Axon Enterprise at Sohn 2026
TL;DR
- Rahul Kishore, founder of Epicenter, pitches Axon Enterprise (AXON, a $30B public company) as “the company building AI agents for police officers,” believing the stock can double over three years. He expects revenue to nearly triple and profits to grow even more, assuming no multiple appreciation from current levels—with additional upside if the market recognizes Axon as an AI winner.
- The installed base is the foundation: by Epicenter’s estimates, over 85% of U.S. police departments use Axon hardware, typically all of it, on 7–10-year contracts with 95% recurring revenue—yet Axon captures only 1% of police budgets. TASER is a $1B business growing over 20%; body cameras are a $1.7B business growing over 40%. Founded in 1993 and public since 2001, Axon has grown about 30% per year for 2.5 decades, grown revenue 4× over five years alongside EBITDA growth, and still grows over 34% annually.
- Draft One, an AI agent that writes police reports compliant with local laws and regulations from body-camera footage, saves 67% of desk time—about 11 hours a week per officer—in a sector where headcount is not growing. Axon Vision extends this: monitoring city CCTV, deploying drones to incidents, and feeding 911 calls to police, ambulance, and fire responders en route in real time.
- The moat is the data flywheel: over 1M officers wearing body cameras, over 1M CCTV cameras monitored, 911 handling for about 30% of the U.S. population, and 200 petabytes of stored video—“about 40× larger than the entire size of Netflix.” Focused proprietary models are cheap to run, and the AI business carries over 85% gross margins.
- The AI Era plan monetizes this at $570 per officer per month—a 5× increase over the standard plan—and did $750M in bookings in its first full year, from essentially zero the year before. Kishore says it is growing faster at this stage than either TASER or the body camera did and will become a much larger business over time than either existing platform.
- The setup: Axon trades at one of its lowest valuations of the past decade, “caught up in the SaaS-pocalypse”—but Kishore argues it is an exception. Most software CEOs, “if they were truly honest,” would say AI’s invention is “terrible” for them; for Axon, it has transformed customer value, the business model, and long-term growth.
Digest · the substance, structured for research
1. AI as employee, not chat box
- Kishore’s opening frame: over 1 billion people use AI weekly “just as a thought partner... a mystical chat box,” but for Epicenter it is “a full-fledged employee” executing autonomously—demonstrated by the firm’s own agent, Eve, introducing itself.
- The thesis filter is that “the companies who enable this for their customers will create immense value”; his pick is Axon Enterprise (AXON), a $30B public company building AI agents for police officers.
2. A 25-year compounder with 85% penetration and 1% monetization
- Founded in 1993 and public since 2001, Axon has grown about 30% per year for 2.5 decades. Revenue grew 4× over the past five years alongside EBITDA growth, and the company is still growing over 34% annually.
- Its two durable core platforms are TASER and body cameras. TASER is now a $1B business growing over 20%; the body-camera business is $1.7B and growing over 40%, with body cameras now “as ubiquitous as donuts and Kevlar vests” for police departments.
- By Epicenter’s estimates, over 85% of U.S. police departments use Axon hardware, typically all of it. The hardware-and-software business has 95% recurring revenue on 7–10-year contracts, yet represents only 1% of police budgets—“the groundwork for the AI era that is just beginning at Axon.”
3. Draft One and the data flywheel
- Draft One auto-generates police reports compliant with local laws and regulations from body-camera footage; officers review, edit if needed, and submit. It saves 67% of desk time—about 11 hours per week—in a sector where the number of officers is not growing.
- Kishore’s Axon Vision scenario has two cars collide at a busy intersection; the agent identifies the serious incident on city CCTV, deploys a drone from a nearby rooftop, and shares incoming bystander 911 calls in real time with police, ambulance, and firefighters en route.
- His data-flywheel case: over 1M officers wear body cameras, over 1M CCTV cameras are monitored, Axon manages 911 calls for about 30% of the U.S. population, and it stores 200 petabytes of video—“about 40× larger than the entire size of Netflix.” The proprietary data supports focused, cheap-to-run models for use cases such as identifying a collision and determining what to do next; Axon’s AI business is over 85% gross margin.
4. $570/officer AI Era plan meets a SaaS-pocalypse multiple
- The AI Era plan is priced at $570 per officer per month—5× the standard plan. After speaking with police chiefs around the country, Kishore said they viewed the ROI as obvious: more productivity from the existing force, more lives saved, higher performance, and, in their view, the only way to achieve it.
- In its first full year, the plan generated $750M in bookings from essentially zero the year before, materially accelerating Axon’s top and bottom lines. Kishore says it is growing faster at this stage than TASER or the body camera did and will become much larger over time than either platform.
- Despite that acceleration, Axon trades at one of its lowest valuations of the past decade, caught up in the “SaaS-pocalypse.” Kishore believes Axon can double over the next three years and expects revenue to nearly triple and profits to grow more than that. Assuming no multiple appreciation, he sees the stock doubling—and probably more if investors recognize Axon as an AI winner transforming its next decade.