LangChain Founder: Agents Are Code, Data and Evals Are King
hwchase17 · x · 2026-08-08
Harrison Chase highlighted key perspectives on building self-improving agents while announcing LangChain's new Managed Deep Agents.
The core argument outlines two major principles for scaling agentic systems:
- Agents are code: Since agents excel at writing code, any task can be framed as an agent writing code to define another agent for that task. The optimal engineering loop involves iteratively writing code to define the harness, testing multiple models/configurations, and evaluating.
- Data, Evals, and Environments are king: Teams need zero friction to kick off their data loops. Out-of-the-box tracing should allow agents to constantly mine signals from traces, convert them into evaluations, and drive improvements via harness engineering or post-training.
Related event: LangChain Launches Managed Deep Agents Public Beta(9 posts)→
More from coding & agent
- AI Era Programming Paradigm Shift: Code Is No Longer the Asset, But a Cache of Intent — mipsytipsy · 2026-08-08
- Blume Sidecar Hits Mac: Monitor and Tune Your Coding Agents in One Place — sull · 2026-08-08
- Claude Code Defaults to Auto Mode Next Month, Blocking 89% of Dangerous Commands — EricBuess · 2026-08-08
- W&B Launches Local Deployment: Run Weave Instances in Just Three Commands — wandb · 2026-08-08
- Beyond Model Supremacy: The Next Layer of AI Agent Tooling — philipkiely · 2026-08-08
- Developer Replaces Siri with Codex as iPhone Voice Assistant — nickbaumann_ · 2026-08-08