Stop Rationing Tokens: Let the Harness Pick the Model - CastAI
Details
Abstract
Token budgets are out of control, and limiting tokens cripples your developers. Cast AI took a different route.
Laurent Gil, co-founder and president of Cast AI, shows how Cast AI built its own open-source coding harness after its coding-agent bill took off. The key idea is to measure cost per task instead of cost per token, and to let the harness choose the right proprietary or open model for each task based on the outcome.
In this talk:
- Why cost per task matters more than cost per token
- How an outcome-aware harness switches models automatically as new ones ship
- Ferment: milestone-based, self-scoring autonomous coding runs that deploy to staging
- Teleport and Studio: remote sandboxes and a shared Kanban board for agent sessions
About the Speaker
Laurent Gil is Co-founder and President of Cast AI, the leading Kubernetes and AI automation platform, used in production by BMW, Akamai, HuggingFace, and many others. He also runs Kimchi, Cast AI’s business unit building an enterprise AI coding platform: autonomous multi-model agents that plan, build, and review code, with spend governance and model control built in.
Previously, Laurent co-founded Zenedge and served as its Chief Product and Business Officer until its acquisition by Oracle in 2018. He was also CEO and co-founder of Viewdle, acquired by Google in 2012, and a founding investor in CUJO AI, acquired by a Charter-Comcast joint venture in 2019.
