Talk Recap : Agentic AI: Beyond the Buzz (LangGraph Studio V2)

Shashikant Jagtap

Interrupt Lang in London*

Following the incredible LangChain Interrupt conference in San Francisco last month, there was an opportunity to catchup in London . Last week, I had the incredible opportunity to speak at the Interrupt Lang* community meetup in London, hosted by the amazing team behind LangChain in London. It was a full house of great people in the Agentic AI space — and a real joy to be part of. The video of the talk yet to be released.

Meetup Link
Demo Video (Langraph Studio V2)
SlideShare Presentation

Talk Overview: Agentic AI – Beyond the Buzz

In my session, I explored what lies beyond the hype of AI agents — diving into the practical, structural, and future-facing aspects of Agentic AI:

Key Themes:

  • What is an AI Agent?
    A deliberately vague slide titled “God Knows” (because let’s face it — everyone defines it differently).

  • Agentic AI Defined
    Beyond chatbots and RAG: autonomous, goal-driven, resilient systems that plan, reason, and interact with the external world.

  • Where It’s Going
    Adoption is rising, particularly in low-risk, high-volume knowledge work tasks. Immediate applications? Think Business Process Automation and Software Development Life Cycle.

  • Agentic Stack & Tools
    Highlighted multiple frameworks in the market — then zoomed in on LangGraph.

Spotlight: LangGraph Studio V2

LangGraph is leading graph-based agent framework for building structured, memory-rich, long-running agents. I mentioned some other frameworks but obviously it’s LangChain meetup so I covered LangGraph

What I covered brief Introduction to LangGraph and it’s Graph based architecture.

One of the highlights of the session was a live demo of the new LangGraph Studio V2 — now a full-fledged web app!

New Features:

  • UI-based input/output tracing

  • Integrated prompt management

  • Real-time LLM call inspection

  • Visual graph flows and human-in-the-loop support

I showed how to build a simple ReAct agent from scratch using LangGraph Studio.

Closing Thoughts

Agentic AI tools and Agent frameworks are maturing fast. The community is growing faster. And we’re just scratching the surface of what’s possible.

Huge thanks to the organizers of Interrupt Lang*, the LangChain London ambassordor (Caheyoon Kim), and everyone who made this event such a vibrant exchange of ideas. Also, shoutout to the other inspiring speakers — it’s exciting to be part of a community shaping the future of AI.

If you’re curious about the talk or want to explore the demo further, check out the video or the slides.

Let’s keep building in Superagentic way.