Ramp built an AI agent harness in six weeks — and LangChain’s CEO says it’s part of a bigger shift
Ramp developed an internal agent harness called Laminar after evaluating existing managed APIs and open-source frameworks. The platform integrates agent execution, sandboxing, observability, and asynchronous workflows. LangChain's CEO praised the design, noting similar efforts in LangChain's Deep Agents project. Both companies' work reflects a growing trend of enterprises building custom agent infrastructure to avoid vendor lock-in and maintain model flexibility.
Key points
- Ramp built Laminar, an in-house AI agent platform, in six weeks to avoid third-party constraints
- Platform provides execution, sandboxing, observability, and asynchronous workflow capabilities
- Trend reflects enterprise preference for custom agent infrastructure over vendor solutions
- LangChain CEO highlighted similar architectural approaches in Deep Agents project
Summary and key points written by AI from the headline and a short excerpt, so they can be incomplete or wrong. The New Stack did the reporting; read the original for the full story.
Read the full story at The New Stack
