
Optimize Agent Costs Without Compromising Quality
Lotte Verheyden
Developer Relations Engineer · Langfuse
Short, practical sessions with the people building and applying AI right now.
Short, practical sessions with the people building and applying AI right now.
Showing 9 sessions

Lotte Verheyden
Developer Relations Engineer · Langfuse

Reduce your AI agent’s running costs with Lotte Verheyden from Langfuse. Learn how to use tracing and evals, run consistent experiments, and choose cost-saving changes while checking their impact on quality.
Lotte VerheydenLangfuse

Arjun Patel
Senior Developer Advocate · Pinecone

Improve your RAG system’s retrieval with Arjun Patel from Pinecone. Explore how to organize your data, choose the right search method, and benchmark retrieval quality for production AI applications.
Arjun PatelPinecone

Mani Srinivasan
Forward Deployed Engineer · Integrations & Developer Tooling · Tavily (Nebius)

Watch a Tavily-powered agent turn web research into a company brief backed by sources. Join Mani Srinivasan and Harsh Kataria for a live demo, code walkthrough, and practical lessons on source quality, latency, and cost.
Mani SrinivasanTavily (Nebius)

Isha Chaturvedi
Former Amazon AGI Researcher. Founding Technical Member · Keenable AI

Build a practical web-research feature with Keenable Agent. Join Isha Chaturvedi from Keenable AI to see how an agent searches the web, reads sources, and returns cited answers. Learn to work with structured outputs, live streaming, and follow-up questions—and evaluate answer quality before using the results in your application.
Isha ChaturvediKeenable AI

Nikhil Nainani
Forward Deployed Engineer · Palantir

Nikhil Nainani shares lessons from four years as a Forward Deployed Engineer at Palantir. Explore customer discovery, ambiguous requirements, technical judgment, and the work of turning business needs into deployed AI systems.
Nikhil NainaniPalantir

Jess Weng
Education Engineer · LangChain

Build agents that can plan, manage context, and delegate work across longer tasks. Jess Weng explains LangChain Deep Agents, including memory, skills, summarization, context offloading, and subagents, and shows how these pieces support practical workflows.
Jess WengLangChain

Brandon Middleton
Head of Education at Replit · Replit

Build and refine an app in Replit, then extend it with reusable skills, integrations, and routines. Brandon Middleton shows how to guide changes, shape the design, connect tools, and handle recurring work beyond the first prompt.
Brandon MiddletonReplit


Agentic workflows need decisions at every step: which tool to use, what to prioritize, and when to ask for help. Jev lets you turn those questions into structured outputs your code can use. In this session, you'll build a small request-routing workflow and learn how to handle uncertainty before it triggers the next action.


Open models are getting dramatically larger and more capable, which makes serving them efficiently a systems problem of its own. This session explores modern inference techniques used to make state-of-the-art open models practical in production, covering caching, kernels, quantization, speculative decoding, and disaggregated inference, with lessons from the recent Kimi K3 launch and its production-scale vLLM integration.
Sujee ManiyamNebius