GenAI solutions with Langchain: Lance Martin on LLMs, agents, evals, and more!

This podcast episode features Lance Martin, a software engineer at Langchain, as he discusses his journey from the realm of computer vision to the world of large language models (LLMs). The hosts, Jennifer Reif and Andreas Kollegger, delve into Langchain's role as an application development framework, simplifying the integration of LLMs and other components while addressing the challenges of productionizing prototypes through the Langchain Expression Language. Lance provides valuable insights into Retrieval Augmented Generation (RAG), emphasizing best practices while advocating for minimal viable chains and the role of LLMs as evaluators in this development process. The episode concludes with recommendations for useful tools and resources, underscoring the importance of collaboration and community in the evolving landscape of AI development.

Outlines

Sign in to continue reading, translating and more.

Continue

Neo4j

Introduction to Lance Martin and his Langchain Journey

Langchain: An LLM Application Development Framework

Langchain's Evolution and the Challenges of DSLs

Deep Dive into RAG: Techniques and Best Practices

Evaluation in RAG: The Importance of Minimal Viable Chains and LLMs as Evaluators

Tools of the Month and Closing Remarks

GenAI solutions with Langchain: Lance Martin on LLMs, agents, evals, and more!

Neo4j

00:00Introduction to Lance Martin and his Langchain Journey

Introduction to Lance Martin and his Langchain Journey

03:04Langchain: An LLM Application Development Framework

Langchain: An LLM Application Development Framework

13:21Langchain's Evolution and the Challenges of DSLs

Langchain's Evolution and the Challenges of DSLs

22:59Deep Dive into RAG: Techniques and Best Practices

Deep Dive into RAG: Techniques and Best Practices

34:40Evaluation in RAG: The Importance of Minimal Viable Chains and LLMs as Evaluators

Evaluation in RAG: The Importance of Minimal Viable Chains and LLMs as Evaluators

42:48Tools of the Month and Closing Remarks

Tools of the Month and Closing Remarks