Recap: Harmonic, LangChain, and the fractious horse

On Wednesday, Max Ruderman (CEO of Harmonic) and Harrison Chase (CEO of LangChain) sat down for a fireside chat in midtown Manhattan. The two led a vibrant discussion of what it takes to build a winning AI-native product in 2025.

A few notable takeaways:

• To translate cutting edge AI capabilities into effective products, you must ride the fractious horse. That is, compress the iterative cycles of development by rapidly building and breaking things instead of taking long cycles of planning and theorizing.
• Composability is core to building a robust agentic ecosystem, and LangGraph provides a lightweight yet opinionated framework for composing processes via graphs. Graphs can then be invoked directly for purpose-built workflows, provided as tools to a higher level agent for arbitrary tasks, or stitched deterministically into other graphs as sub-graphs.
• Thoughtful UX can make or break an AI-based product. LLM and agent driven systems tend to produce far slower experiences than users have come to expect of software in recent years, so the burden falls on the product builder to visually demonstrate the value and breadth of work that the agents are performing.

In attendance were engineers, founders, and investors representing major companies and startups alike – united by an interest in the practical implementation of AI to solve real world problems.

We would like to extend gratitude to the 300 attendees on our RSVP list, as well as the 300+ waitlisted guests who we were unfortunately not able to accommodate for this event. To stay in the loop for our next event, join our mailing list below.

‍