Aaron Linsky of Bridgewater Associates and Nikhil Benesch of turbopuffer explain how permission-aware retrieval over a 100-million-document research vault broke a conventional search setup, and why native filtering over a clustered vector index is needed to keep search fast without sacrificing recall.
Bridgewater Associates' Applied AI and product engineering leads explain how they built the Vault, a governed repository of more than 100 million unstructured documents, and why humans and agents need different things from the same search system. They cover a hand-built benchmark that lifted precision at 10 from about 60% to over 90%, the filtering and scaling choices behind it, and the shift toward proactive agents that bring investors the topics most deserving of their attention.
A hands-on GraphRAG workshop from Neo4j showing how combining vector search with a knowledge graph gives an agent connected context, multi-hop traversal, and explainable answers. It covers the difference between semantic similarity and relationships, LLM-driven entity and relationship extraction under a fixed schema, building an Aura database and Python pipeline, and configuring an Aura agent with similarity search and text-to-Cypher, with a closing check that shows why visible query evidence matters.
George He, head of engineering at LlamaIndex, explores when agents need pre-indexed vector search and when plain file traversal with grep is enough. He contrasts Claude Code's embedding-free approach on code with the harder reality of large, messy, multimodal, permissioned company documents, and proposes a harness of hybrid retrieval, metadata filtering, grep, and reading tools backed by LlamaParse and LiteParse. The talk closes with a live demo of an agent building a cash-flow table from 135 Alphabet filings, grounded in source files.
This talk by Jess Wang from Braintrust delves into the critical importance of evaluations (evals) in AI development, moving beyond subjective "vibes" to data-driven decision-making. She breaks down the four core components of an eval—dataset, task, scoring, and experiments—and illustrates the continuous improvement "eval flywheel" integrating observability. The central experiment compares agentic search against vector search for a coding agent tasked with finding buggy code in Microsoft's TypeScript Go repo. While both methods achieved identical accuracy, vector search proved four times more expensive due to its inability to capture surrounding context within code chunks, leading to excessive, inefficient searches. Wang concludes with practical caveats and best practices for conducting effective LLM evals.
Dylan Couzon, developer relations engineer at Qdrant, argues that frontier-class models now run on hardware people can own, but the memory that makes an assistant personal still lives in someone else's cloud. He covers why owning inference gives autonomy while owning memory gives continuity, the CPU, RAM, and disk model for LLMs, why retrieval beats stuffing everything into the prompt, and a live offline demo of a drone building searchable memory with Qdrant Edge, plus opt-in shared memory.