← back
3 ingredients for building reliable enterprise agents - Harrison Chase, LangChain/LangGraph
Takeaway
Reliable enterprise agents come from picking high-value tasks, mixing workflows with agents, and using observability to shrink uncertainty rather than chasing pure autonomy.
Summary
- Enterprise agent adoption follows: value when right × probability of success − cost when wrong, minus run cost.
- High-value verticals (Harvey for legal, finance research) and ambient long-running agents (deep research, background coding) increase upside.
- Reliability comes from blending workflows and agents, not choosing one — LangGraph spans the spectrum so deterministic steps can constrain LLM freedom.
- Observability/eval tools like LangSmith reduce stakeholder error bars by exposing each LLM call and benchmarking against ground truth.
- Lowering cost-when-wrong via human-in-the-loop and tight scopes is as important as raising capability.
langgraphenterprise-agentsobservability
Original description
It's easy to build a prototype of an agent, but hard to put an agent in production - especially in an enterprise setting. In this section, will talk about three ingredients for building reliable agents in the enterprise. About Harrison Chase Harrison Chase is the CEO and co-founder of LangChain, a company formed around the popular open source Python/Typescript packages. The goal of LangChain is to make it as easy as possible to use LLMs to develop context-aware reasoning applications. Prior to starting LangChain, he led the ML team at Robust Intelligence (an MLOps company focused on testing and validation of machine learning models), led the entity linking team at Kensho (a fintech startup), and studied stats and CS at Harvard. Recorded at the AI Engineer World's Fair in San Francisco. Stay up to date on our upcoming events and content by joining our newsletter here: https://www.ai.engineer/newsletter