LangChain and TypeSafe AI Host Offline Tech Event on Decision Models and Agent Evaluation

iconKuCoinFlash
Share
AI summary iconSummary
LangChain and TypeSafe AI will host an in-person technical event on November 5 in San Francisco’s SoMa, focusing on decision models and agent evaluation. The event will not be live-streamed and is open exclusively to registered attendees. Speakers include Sydney Runkle, Jev, and Allie Laabs, who will cover topics such as LangGraph and agent testing with low-cost models. Seán Roche will present a model comparison using on-chain data to measure accuracy, consistency, and cost. The session runs from 18:00 to 20:30. LangChain’s tools serve over 7,000 clients, including LinkedIn and Workday. The event takes place amid mixed sentiment reflected in the Fear & Greed Index.
ME AI event: LangChain and TypeSafe AI will host an in-person technical meetup on November 5 in SoMa, San Francisco, focusing on building decision models and their application in agent evaluation. The event will not be live-streamed or recorded and is open only to registered and approved attendees. The three speakers and their topics are as follows: Sydney Runkle, Product Manager at LangChain’s open-source team, will cover decision models and LangGraph, noting that most agents delegate each decision to cutting-edge large models, resulting in slow speeds and high costs; Jev from TypeSafe AI can provide calibrated probabilities for type-specific agent states and introduces a hybrid pattern of “default to low-cost models, invoke cutting-edge models only during anomalies.” Allie Laabs, Founding Developer Evangelist at TypeSafe AI, advocates understanding such models as “System One Models.” LangChain Education Engineer Seán Roche will share real-world testing of Jev as a judge: the team replayed identical agent runs through Deep Agents and LangSmith to Jev, GPT-5.6 Luna, GPT-5.6 Terra, and Claude Sonnet 4.6, comparing accuracy, consistency, latency, and cost. The agenda begins at 18:00 with welcome and refreshments and concludes at 20:30. Hosted by LangChain, which provides a model-agnostic agent framework and LangSmith for testing, deployment, and monitoring—with over 7,000 customers including Nvidia, Bridgewater, LinkedIn, Workday, Harvey, and Rippling. (Source: InFoQ)
Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information. Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.