Historically, unstructured data has dominated the spotlight in AI, while the mission-critical structured data that drives most enterprise workflows has remained under-leveraged, with few proven recipes for AI workloads. Today, we’re changing that by fully open-sourcing Contextual-SQL, a state-of-the-art Text-to-SQL pipeline which ranks highly on the BIRD benchmark and you can run entirely on-prem. A surprisingly simple pipeline delivers these results by leaning on two core ideas: 📖 Context beats parameters DDL → mSchema (table column comments) → mSchema one few-shot example lifts execution accuracy from 54.7 % to 62.5 %. Before reaching for a larger model, enrich your schema docs and drop in a golden demo query. 📈 Scale at inference Spin up 1000 diverse SQL candidates in parallel, filter invalid queries with a fast sqlite3 check, then rank what’s left using a lightweight reward model built on the same Qwen base plus log-prob confidence. That single trick bumps pass@1 to ~73% -- cheaper and cleaner than fine-tuning. The whole flow is just five step: generate → filter → rank → pick → run, and lives on GitHub. Fork it, point it at your schema and ship a private text-to-SQL solution. For a deeper dive, code, and benchmarks, see Sheshansh’s thread and the full blog post below.

🚀 Just had a deep, wide-ranging convo on the Five Year Frontier pod with @DanielDarling about the future of AI work in the enterprise. Some highlights 👇 1/ Why generic LLMs don’t cut it for real-world enterprise use 2/ How Grounded Language Models (GLMs) eliminate hallucinations 3/ Teaching AI your org’s tribal knowledge 4/ Agents that grow with you over time 5/ How RAG 2.0 our new reranker bring control precision to enterprise AI 6/ What a “thin, high-leverage” AI-native org might look like in 5 years @ContextualAI We’re building toward a world where AI understands your workflows, your knowledge base, and your org structure. That’s the new operating system of modern work. 🎧 Listen here: youtube.com/watch?v=d85Kf_YM…