Why Custom Software Might Kill Vertical AI Startups
Tenex and swyx debate whether vertical AI startups can survive custom enterprise software and cheap frontier models.
AI in NYC: How Tenex Builds an AI-Native Engineering Team — Alex Lieberman & CJ Hess
With swyx, Alex Lieberman, CJ Hess · Sunday, October 25, 2026
Alex Lieberman and CJ Hess of Tenex join swyx at AI Engineer New York to explain how they run an AI-native engineering consultancy that pays software engineers based on story points and output rather than billable hours. They detail their MetaHarness development workflow that commits agent plans and decisions into Git, outline the enterprise diffusion gap where only 5% of non-engineering agents reach production, and debate whether vertical AI startups and mid-tier SaaS can survive custom internal rebuilds.
Key takeaways
- Tenex co-founders Alex Lieberman and CJ Hess argue vertical AI startups face extinction as buyers demand custom software built for their exact workflows. Read more →
- Only 5% of non-engineering enterprise AI agents make it into production, stalled by fragmented human workflows rather than model limitations. Read more →
- 90% of large enterprises have purchased enterprise licenses for frontier models like Claude or ChatGPT, but buying seats does not mean workflows change. Read more →
- Tenex pays software engineers strictly by completed story points instead of billable hours, linking compensation directly to delivered code. Read more →
4 articles from this episode
Why 95% of Enterprise AI Agents Fail to Reach Production
Tenex founders Alex Lieberman and CJ Hess explain why enterprise AI agents stall on messy internal operations and missing RBAC security controls.
Alex Lieberman on the Enterprise AI Adoption Gap
Alex Lieberman tracks enterprise AI adoption across four milestones, exposing why only 5% of companies put non-engineering agents into production.
Tenex Pays Engineers on Story Points Instead of Hours
Tenex ditched hourly billing to pay engineers for story points. Top performers now make over $1M while output variance hits 4x.
More Latent Space episodes
- The Programmable Data Center — Evan Conrad, SF ComputeOct 25 · 6 articles
- Robots That Predict the Future: Dyna-2 & Physical AI — Jason Ma, Dyna RoboticsOct 18 · 4 articles
- High Taste Testers: The AI Tools We Keep Coming Back ToOct 18 · 5 articles
- From Wet Lab to RL Env: 10,000 Drug-Discovery Campaigns in a Weekend — Ian Quigley, Leash BioOct 18 · 6 articles