RavenDB Quill: AI Agents on Your SQL Database in Weeks

RavenDB launched Quill on September 8, a context layer that connects AI agents to existing SQL databases without migration, aiming to reduce production deployment from 18–24 months to weeks. Quill supports PostgreSQL, SQL Server, and MySQL, uses change data capture for real-time sync, and treats AI as untrusted by default with access controls independent of database permissions. The launch comes as research from Digital Applied shows 79% of enterprises have AI agents but only 11% run them in production, and Gartner projects 40% of enterprise applications will integrate task-specific AI agents by end of 2026.

The demo-to-production gap is enterprise AI’s least glamorous problem. You can have an AI agent querying your SQL database in an afternoon. Getting that same agent running in production — with real-time data sync, access governance, and zero changes to the system of record — routinely takes 18 to 24 months of internal engineering work. RavenDB is betting that timeline is unnecessary. On September 8, the company launched Quill , a context layer that sits between AI agents and existing SQL databases, aiming to compress that multi-year project to weeks. A Layer, Not a Migration Quill connects to your existing PostgreSQL, SQL Server, or MySQL database without touching it. The source database stays authoritative; Quill builds a governed, AI-ready context layer on top, using change data capture to stay synchronized with live data. It handles vector embeddings automatically, includes retrieval-augmented generation RAG and semantic search, and uses the Model Context Protocol MCP https://modelcontextprotocol.io/ as the communication layer between agents and your data. No separate vector database. No custom data pipeline. Quill positions itself as the data pipeline. “Anyone can stand up an AI demo in an afternoon, but getting that demo into production with data pipelines, semantic search, security, governance — all the plumbing a small proof of concept doesn’t need until it has to run at scale — is the hard part. Quill exists because we’d rather hand teams that plumbing already assembled than watch them rebuild the same project after project.” Oren Eini, Founder and CEO, RavenDB Zero Trust Toward AI The governance model is where Quill does something interesting. Rather than trusting the AI model with broad database access and hoping the prompting holds, Quill treats the AI as untrusted by default. Organizations configure access controls at the Quill layer — independently of whatever permissions exist at the database level. A healthcare deployment can give an agent full access to appointment scheduling data while completely blocking prescription records. The agent never sees what it is not allowed to see, regardless of how a prompt is crafted. This matters in September 2026, when EU AI Act enforcement is actively rolling out and SOC 2 auditors are increasingly scrutinizing AI agent data access patterns. Pointing an LLM at a raw database and relying on system prompts for access control is no longer a tenable enterprise architecture. The Production Gap Is Real Quill’s pitch lands in context. According to research from Digital Applied https://www.digitalapplied.com/blog/ai-agent-scaling-gap-march-2026-pilot-to-production , 79% of enterprises now have AI agents in some form, but only 11% are running them in actual production — a 68-percentage-point gap between “piloted” and “shipped.” Integration complexity with legacy systems is the single largest driver of that gap. Gartner projects that 40% of enterprise applications will integrate task-specific AI agents by the end of 2026, up from under 5% in 2025. Something has to bridge that distance. “With Quill, the plumbing was already there, so we spent our time building the actual feature.” Hagay Albo, CEO, Albos Technologies Deployment and Availability Quill-powered agents deploy across web chat, WhatsApp, Telegram, Slack, and Discord. The system is model-agnostic — it works with Anthropic, OpenAI, Google, or any self-hosted model — and supports cloud-hosted or on-premises deployment for regulatory compliance. RavenDB has not disclosed pricing. Quill is available now https://www.globenewswire.com/news-release/2026/09/08/3357703/0/en/ravendb-launches-quill-to-bring-production-ai-agents-to-enterprise-sql-systems-no-migration-required.html for PostgreSQL, SQL Server, and MySQL, with additional database support planned. The real test will be how the architecture holds up against actual enterprise complexity — sprawling access requirements, legacy schemas, and active compliance audits. For teams that have watched AI agent projects stall in the infrastructure phase for the third time, it is at least a more compelling starting point than building from scratch again.