Why attend:
Most “RAG” demos crumble on a real-world scale. In this hands-on session, Andriy Burkov will take Domyn’s FinReflectKG dataset—a graph with ~17.5 million factual triplets extracted from a decade of S&P 500 10-Ks—and show how to turn it into a production-ready knowledge base in MindsDB. Then he’ll build an agent that composes KB queries on the fly to produce grounded, explainable answers to deep analytics questions.
You’ll see end-to-end patterns for going from a huge 10-k filings archive to fast, precise, source-linked answers—without brittle prompt spaghetti.
What you'll learn:
Load and organize a multi-million-triplet dataset into a MindsDB Knowledge Base designed for speed, recall, and verifiability.
Structure and execute dynamic KB queries that pull exactly the facts you need.
Practical tips for grounding, evaluation, latency at millions-of-facts scale.
Who should attend:
Data/ML engineers, analytics leaders, AI product builders, enterprise architects—anyone who needs trustworthy answers from large corporate text archives, not just summaries.
Speakers
Andriy Burkov, Ph.D.
Renowned AI/ML expert and MindsDB advisor, bestselling author.
Andriy is a best-selling author and 20‑year machine‑learning veteran who has led dozens of AI projects at Gartner, Fujitsu and TalentNeuron.
Alejandro Cantu
AI Product Manager, MindsDB
Alejandro drives product strategy that helps organizations connect AI to existing data for real‑time insights, drawing on prior experience at Google and as a tech founder.
