- Joined Cognition
- Jul 2025
- How they joined
- With Windsurf
- Before Cognition
- MedalliaSenior Sales Director, Enterprise Sales
- Status
- Current
- Marked one year since Cognition acquired Windsurf (2025-07-15) calling it the 'wildest 72 hours' of his career, with recruiters and customers calling and nobody sleeping; credits Scott Wu and Russell Kaplan for seeing that 'the best engineering org and the best GTM org in the AI Space belonged together'.Jul 15, 2026One year ago today, Cognition acquired Windsurf. Wildest 72 hours of my career, no question. Recruiters calling me, my friends, and our team. Customers calling to figure out what was happening. Nobody sleeping. It felt like the ground was moving under us in real time. Then Scott Wu and Russell Kaplan stepped in. They saw something the rest of the market hadn't caught onto yet, that the best engineering org and the best GTM org in the AI Space belonged together. A year later, watching what Devin has become, it's clear how far ahead of the moment they actually were. Grateful I got to be part
- Joined Windsurf (Codeium) as GTM - Strategic Account Executive in Mar 2025 after leaving Medallia in Feb 2025; Cognition acquired Windsurf in July 2025 and he continued in the same strategic AE role at Cognition from Jul 2025.Mar 2025
- Posted about Wells Fargo Technology engineers spending four days at Devin Co:Lab in San Francisco, saying he has 'been lucky to work alongside this team for a while now'.Oct 29, 2026A line we use a lot at Cognition: it's fun being on the Avengers. Having worked so closely with the Wells Fargo team, they've felt like part of the squad from day one. Co:Lab was Avengers, assemble. Wells Fargo Technology engineers joined our team at Devin Co:Lab in San Francisco for four days of building together. Then they took real problems from their day jobs and solved them with Devin. I've been lucky to work alongside this team for a while now. Being in the room for this, and feeling that energy firsthand, made it obvious. Something special has been building here for a long time, and
- Current Cognition role is described as 'Helping enterprises adopt Devin, the world's first AI software engineer', based in Austin, Texas.Jul 2025
- His farewell post (2025-03-10) said he was 'thrilled to begin my next chapter at Codeium!' and thanked Meredith Siewert (McCarty), Sebastian Cañizares, Paxton Duff, Chris Li, Dave Frankel, and Sandeep Shah for 'your help getting here', which names the people who helped bring him from Medallia into Codeium/Windsurf.Mar 10, 2025
- Wrote that in 2024 'Enterprise demand was overwhelming. Devin wasn't ready yet' and that 18 months later Devin is embedded 'from tech-native companies to the most regulated institutions in the world'.Aug 19, 20262024: Devin launches. Enterprise demand was overwhelming. Devin wasn't ready yet. 18 months later, that's completely flipped. I've seen Devin get embedded inside enterprises and reshape how they operate, from tech-native companies to the most regulated institutions in the world. The deck is getting reshuffled. Everywhere. You're going to hear and see more about Devin. The teams going for it all aren't waiting around.
- Described two weeks with seven flights across the US meeting lead engineers and C-suite executives wrestling with 'what does the workforce look like once you embed an AI engineer into it', showing a field-heavy enterprise selling motion.Jul 24, 2026Two weeks. Seven flights across the country. East Coast, West Coast, Southeast, and everywhere in between. In that stretch I sat with lead engineers and I sat with C-suite executives, all wrestling with the same question: what does the workforce look like once you embed an AI engineer into it. You always hear it from mentors further down the road: “I wish I’d known how special that time was back then.” Somewhere in the middle of all this travel, it hit me. This is one of those stretches you don’t recognize as special until years later. I’m trying my best to notice it while I’m still in it,
- Former Medallia colleague Jennifer Ortegon posted 'Medallia ties go deep! So fun getting to work with Canyon Manske and Jake Villas again at Windsurf!' in July 2025, showing a Medallia-to-Windsurf sales cohort moved together.Jul 10, 2025
- Most recent post (2026-11-03) shared Cognition's 'A Practical Guide to Security in November 2026', arguing the developer laptop is 'the front door to prod'.Nov 3, 2026$6.34 billion walked out of Binance. It started with one engineer's laptop. A fake job interview. A tampered installer. A deployment token on the machine. In the same two weeks, a malicious Claude Code extension for Cursor auto-updated onto developer machines and tried to backdoor them. The developer laptop is now the front door to prod. It runs packages, extensions, MCP servers, and AI agents that run commands on their own. The fix isn't fancy. Get secrets off laptops. Lock down extensions. Run your agents somewhere you control. Must read. https://lnkd.in/gB9GgypF
- Promoted Cognition's productive-engineering-hours estimator and the guarantee to 'fund your usage up to $10M' if Devin delivers less value than customers pay, a value-guarantee argument useful in enterprise deals.Jun 4, 2026Tokens burned. Lines generated. Sessions completed. None of it tells you if the work was actually useful. We built an estimator that evaluates every Devin session, was the output productive, and how long would a human engineer have taken to do the same work? Validated against real engineers' time estimates in enterprise codebases. It's not an ROI calculator. True ROI is harder. But productive engineering hours is a real signal, and a meaningful place to start. When we ran our historical data through it, we felt confident enough to back it with a guarantee: if Devin delivers less engineering
- Commented on the xAI-Cursor deal that 'the strongest independent players' are accelerating amid consolidation, positioning Cognition as an independent winner.Apr 23, 2026The xAI and Cursor deal has a lot of founders asking if AI startups need to sell or die. Our experience suggests the opposite: as more companies consolidate, the strongest independent players—those with clear differentiation, focus, and velocity—are accelerating.
- Spent Aug 2021-Feb 2025 as Senior Sales Director, Enterprise Sales at Medallia (Austin), the role immediately before Windsurf.Aug 2021
Helping enterprises adopt Devin, the world’s first AI software engineer.
Cognition acquired Windsurf in July 2025
Create and deliver a presentation to maximize and improve customer satisfaction survey responses. Act as strategic advocate to ensure clients are successful and grasp the value of their investment with Oracle. Provide best practice and knowledge on the Oracle CX products and solutions.
CSM's are responsible for leveraging Oracle resources, programs, and information in order to ensure the value of our customer's Oracle Cloud service. Customer satisfaction and retention are the two primary goals of my role. Ensure value through tracking key performance indicators, quarterly health checks, and annual business reviews.
- Performed administrative and clerical assistant to operations and recruiting staff - Formed prospective profiles for target recruits including highlights, social media profiles, logged calls, and communication exchanges. - Performed roles in prospective player identification, talent evaluation, customer service, and marketing as part of the team's recruiting strategy - Assist with Game Day preparation and operation - Reach out and engage with media members and outlets to confirm recruit’s performance or attendance.
One year ago today, Cognition acquired Windsurf. Wildest 72 hours of my career, no question. Recruiters calling me, my friends, and our team. Customers calling to figure out what was happening. Nobody sleeping. It felt like the ground was moving under us in real time. Then Scott Wu and Russell Kaplan stepped in. They saw something the rest of the market hadn't caught onto yet, that the best engineering org and the best GTM org in the AI Space belonged together. A year later, watching what Devin has become, it's clear how far ahead of the moment they actually were. Grateful I got to be part
A line we use a lot at Cognition: it's fun being on the Avengers. Having worked so closely with the Wells Fargo team, they've felt like part of the squad from day one. Co:Lab was Avengers, assemble. Wells Fargo Technology engineers joined our team at Devin Co:Lab in San Francisco for four days of building together. Then they took real problems from their day jobs and solved them with Devin. I've been lucky to work alongside this team for a while now. Being in the room for this, and feeling that energy firsthand, made it obvious. Something special has been building here for a long time, and
2024: Devin launches. Enterprise demand was overwhelming. Devin wasn't ready yet. 18 months later, that's completely flipped. I've seen Devin get embedded inside enterprises and reshape how they operate, from tech-native companies to the most regulated institutions in the world. The deck is getting reshuffled. Everywhere. You're going to hear and see more about Devin. The teams going for it all aren't waiting around.
Most AI coding tools get expensive as you scale. Devin just got cheaper and not by cutting corners. Devin Fusion now leads FrontierCode 1.1 Extended. Best score on the benchmark, at $0.60 a task. Two reasons this works: Model independence. Devin doesn't bet on one frontier model. It routes each step to whatever model is actually best for that step — and gets cheaper every time any of them improves. A harness built to not waste tokens. Fewer redundant tool calls. More context pulled from cache instead of reprocessed from scratch. Net result: more capability, running on fewer tokens, for le
Two weeks. Seven flights across the country. East Coast, West Coast, Southeast, and everywhere in between. In that stretch I sat with lead engineers and I sat with C-suite executives, all wrestling with the same question: what does the workforce look like once you embed an AI engineer into it. You always hear it from mentors further down the road: “I wish I’d known how special that time was back then.” Somewhere in the middle of all this travel, it hit me. This is one of those stretches you don’t recognize as special until years later. I’m trying my best to notice it while I’m still in it,
Tokens burned. Lines generated. Sessions completed. None of it tells you if the work was actually useful. We built an estimator that evaluates every Devin session, was the output productive, and how long would a human engineer have taken to do the same work? Validated against real engineers' time estimates in enterprise codebases. It's not an ROI calculator. True ROI is harder. But productive engineering hours is a real signal, and a meaningful place to start. When we ran our historical data through it, we felt confident enough to back it with a guarantee: if Devin delivers less engineering
March 2024: The internet roasted us. 18 months later: The lesson isn't that the haters were wrong. The lesson is what Scott saw that the critics didn't. Models compound. Products compound. Customer feedback compounds. The team that ships early and absorbs the punch gets 18 months of real-world signal the team waiting for "the right launch" never sees. By the time anyone else was ready to compete, we'd already learned what enterprises actually need. We'd built the connective tissue. We'd shipped real deployments under real production load. 2024 wasn't the embarrassing part. It was the star
$6.34 billion walked out of Binance. It started with one engineer's laptop. A fake job interview. A tampered installer. A deployment token on the machine. In the same two weeks, a malicious Claude Code extension for Cursor auto-updated onto developer machines and tried to backdoor them. The developer laptop is now the front door to prod. It runs packages, extensions, MCP servers, and AI agents that run commands on their own. The fix isn't fancy. Get secrets off laptops. Lock down extensions. Run your agents somewhere you control. Must read. https://lnkd.in/gB9GgypF
This is the most excited I've been for a launch in recent time. Devin Oncall is live. The 3am page. The scramble through dashboards and logs. The alert that turns out to be noise. Again and again. Now Devin takes that alert first. It finds the root cause, pulls in the right person if needed, and opens a PR with the fix. And it's already live in production, handling thousands of alerts a day for some of the world's largest companies. On-call has always been the tax engineers pay for shipping software. Devin just started paying it for them.
The xAI and Cursor deal has a lot of founders asking if AI startups need to sell or die. Our experience suggests the opposite: as more companies consolidate, the strongest independent players—those with clear differentiation, focus, and velocity—are accelerating.
"It feels like magic at first. It isn’t. It’s a new operating model for software." Well said Alexandra Mysak. Very humbling to see how quickly Devin has moved from edge case to enterprise must.
Tuesday brought critical PHP, Node.js, and Linux kernel vulnerabilities. One of our financial services customers hit this directly: a client-facing trading platform on unsupported PHP, no tests, no in-house expertise. Using Devin, they rebuilt the runtime and generated 1,000+ tests, back in production in six hours, zero client impact. Come hear Alex Stamos this week. Well worth your time.
