Ramp explains why it built its own agent harness, LangChain CEO weighs in
hwchase17 · x · 2026-10-21
Ramp published a deep dive on why it built its own cloud-native, model-agnostic agent harness instead of using off-the-shelf options, citing model/provider independence, co-building with product teams, and owning its observability stack. LangChain CEO Harrison Chase endorsed the analysis: many agents look alike internally, harness responsibility grows well beyond a while loop in production, and "sandbox-as-a-tool" plus non-blocking async agents are the direction deepagents is also taking.
More from coding & agent
- RoboARC shows robots get no free retries—unlike coding agents that can just rerun tests — ManlingLi_ · 2026-10-21
- Anthropic open-sources 11 Claude plugins for sales, legal, finance and more knowledge work — mhdfaran · 2026-10-21
- Pixlet display turns Claude agent status into a physical dashboard — whurley · 2026-10-21
- Ding CAD Goes Down, Revealing Its User Base Is Far Bigger Than Its Creator Expected — yacineMTB · 2026-10-21
- If your evals pass 100% of the time, they've stopped telling you anything — HamelHusain · 2026-10-21
- Hamel Husain ranks AI eval methods: engineers can't replace human experts for subjective agent output — HamelHusain · 2026-10-21