The past two months have been an intense stretch of shipping at TableQL. Internally, we've been calling it Perpetual Launch Week. For thirty days straight, we shipped something new every single day.
Behind all of that work has been one clear goal: making TableQL more enterprise-ready, more governable, and more useful to companies wherever they are on their AI maturity curve.
That meant improving the foundation underneath Vera, while also making the product easier to adopt, observe, trust, and extend. Over the past month, we shipped a redesigned Ontology layer for governing business context as version-controlled files optimized for AI locality, an MCP server for Vera, a new Observability page for usage tracking and quality evals, an Ontology onboarding mode to help teams build from scratch, and Citations with Data Lineage.
Below, we'll walk through the five launches that moved the needle most.
Ontology 3.0
Ontology 3.0 is a complete overhaul of our previous ontology graph, merged with our original Context Library. The result: TableQL can now ingest anything beyond data models and semantic layers, including files, dashboards, executables, even agent configs, and structure it all in a lightweight semantic modeling format optimized for AI consumption.
The governance story is just as important. Vera can suggest changes to the Ontology based on usage patterns, and data admins can approve or reject them. Everything is version-controlled, diffed, and subject to proper review: the same best practices software teams apply to code, now applied to your business context. And because everything lives as files and config, it's git-backed. Your ontology goes wherever you go.
The redesigned Ontology is now available to all customers. Read more here.
Vera MCP Server
We launched Vera as an MCP server so your teams can connect to it headlessly, preserving whatever interface your business already runs on, whether that's Claude, ChatGPT, Copilot, an in-house LLM assistant, or a coding tool. Vera plugs into the workflow, not the other way around.
Connect Vera at
app.textql.com/mcp. Read more here.
Observability Page
Observability is the monitoring plane for both usage and evaluation quality. As an administrator, you can track active users, ACU (Agentic Compute Unit) consumption, and connector activity alongside sentiment, response lengths, failed executions, cost per query, and LLM-judged missing context. Any pattern you find can be drilled into with full chat history for diagnosis.
And these insights don't have to live in TableQL. If your team already runs a monitoring stack, you can pipe everything into it via OpenTelemetry-compatible collectors.
Observability is available now. Find it in the bottom-left of your navigation bar. Read more here.
Ontology Onboarding
Most teams know they want an ontology, they just don't know where to start. Onboarding Mode solves the blank-slate problem: a guided, step-by-step flow that structures your context the way you want it. Prefer to let Vera drive? She can explore your connectors and ask the right questions to design your ontology for you.
Either way, you get a faster path to value from day one, and a repeatable process for onboarding new teams over time, whether that's Finance, Legal, or any business unit that comes next.
To enable Ontology Onboarding Mode, click the + button and select Methodology in a new Thread. Read more here.
Citations
Citations in Vera now appear inline, familiar to anyone who's used a modern AI assistant. But what sets them apart is full data lineage: every figure shows exactly where it was sourced from and whether it underwent any transformation on the way to its final state. No black boxes. You can trace any number back to its origin and debug with precision.
Toggle Citations on in Settings. Read more here.
The Full Roundup
Those five were just the headliners. Here's everything else we shipped:
- TableQL Desktop App: Vera, now on your desktop. A native app experience for teams who want Vera outside the browser.
- Connector Explorer Page: A per-connector view into your data sources. Browse the schema, check usage, and see exactly how each connector is being queried by Vera.
- Vera Forms: Structured input flows that let Vera collect exactly what it needs before running an analysis, reducing back-and-forth and improving output quality.
- Dynamic Visualizations: For one-off analyses, Vera now generates charts and visuals that adapt to how you interact with your data, not just static outputs.
- OAuth & SSO Support: Enterprise-grade authentication, out of the box. Connect your existing identity provider and manage Vera access the same way you manage everything else.
- OpenTelemetry: Pipe Vera's usage and performance data into your existing monitoring stack via OpenTelemetry-compatible collectors. Your observability setup, extended to Vera.
- Vera Claude Plugin: Access
/anadirectly inside Claude. Bring your governed business context into your AI workflows without switching tools. - First-Class Connector Support: SAP HANA, Firebolt, Dremio, Oracle, kdb+, Exasol, Attio
Try These New Capabilities Today
If you're already a TableQL customer, these features are live. Open Vera and start exploring. If you're not yet a customer, request a demo or start a free trial.
Follow our newsletter for product updates, and TableQL on LinkedIn for daily release announcements.
