Key Takeaways
- Tenex compensates AI engineers by output (story points) instead of hours.
- This model incentivizes engineers to leverage AI for 10x productivity.
- Engineers are projected to earn over $1 million annually through this system.
- Tenex achieves rapid prototyping, delivering complex systems in weeks.
- Human capital, specifically hiring exceptional engineers, limits Tenex's growth.
- Advanced AI interview questions probe understanding of "controlling entropy."
Deep Dive
- Tenex co-founders Alex Lieberman and Arman Hezarkani developed an output-based compensation model for AI engineers.
- This model, based on story points rather than hourly billing, aims to align incentives for AI tool adoption and maximize throughput.
- The concept emerged after Hezarkani's previous company achieved a 10x increase in software output despite a 90% engineering team downsizing by going AI-first.
- To ensure quality and prevent system gaming, Tenex hires engineers motivated by long-term client relationships or a passion for coding.
- Engineers are compensated by story points, while technical strategists are incentivized by Net Revenue Retention (NRR) and client retention.
- The company anticipates multiple engineers will earn over $1 million annually via this story point compensation model.
- Tenex developed a retail computer vision system with heat maps and theft detection, prototyping in two weeks for work that previously took quarters.
- The company built Snapback Sports' mobile trivia app in one month, which achieved a global App Store ranking of 20th.
- An AI health coach app prototype was developed in four hours for a prospect, dramatically accelerating the sales motion.
- Tenex identifies human capital, specifically hiring exceptional AI engineers quickly, as its primary growth bottleneck.
- Despite being an AI-first company, the challenge lies in finding individuals who can effectively harness AI's leverage.
- While Tenex has ambitions to build proprietary technology, recruiting enough skilled talent remains the immediate constraint.
- Tenex employs "unreasonably difficult" take-home interviews, with approximately 50% of candidates not completing the challenge.
- The interview process is short, involving two calls, a take-home review, and one or two final meetings, potentially completable in a week.
- A signature question asks about the primary bottleneck in building an AI to replace an engineer, revealing advanced insights into AI agent challenges.
- Discussions at the conference highlight trending keywords like "context engineering" and the controversial "MCP" acronym.
- Speakers debate whether new acronyms, such as MCP, represent significant advancements or are used to attract investment.
- One perspective views new terminology as a natural sociological behavior for groups to communicate, with ideas and language recycled over time.