Skip to main content

Retool deepens Databricks partnership: Lakebase and Agent Bricks

Blog article hero image
Connie Wu
Connie Wu
VP of Global Partnerships, Retool

Last year, we were honored to be named the 2025 Databricks Emerging Partner of the Year, in recognition of our shared commitment to help enterprises turn Databricks data into action. A lot has changed since then. The foundation making that possible keeps getting stronger. That's what has us excited to be back at the Data + AI Summit this week, and to share what's next for our partnership.

From insight to action

Across financial services, retail, consumer goods, healthcare, and manufacturing, the same question keeps coming up. You have the data, you have the models, so why is putting that insight to work still the hardest part? That gap between knowing and doing is where value gets lost. Closing it is exactly where we are focusing our partnership, and as launch partners for Lakebase and Agent Bricks, we're closer than ever.

"The future of enterprise AI is grounded in trust and transparency. Retool has become a critical part of how Databricks customers put AI to work across their enterprise. By expanding our partnership to include Lakebase and Agent Bricks, we've unlocked new ways for teams to turn insight into action without sacrificing governance." — Shanku Niyogi, SVP of Product, Databricks

Lakebase Launch Partner

In February, Databricks Lakebase became generally available: a fully managed Postgres database that lives inside the same platform where your analytical data already sits. We are a launch partner because we'd been building on it ourselves and believe this is where operational databases are headed. Separating storage from compute is the architecture that makes scale manageable and Databricks is leading the way.

This foundation is already running in production across more than one million serverless Postgres databases inside Retool. Which is what made the Retool Lakebase connector possible. Now your operations teams can resolve service tickets, manage inventory, and process orders against the same data your analysts are already using—without provisioning anything.

The operational layer around AI is becoming just as important as the model layer. With the Retool Lakebase connector, our teams can respond to fleet and operations data in real time and scale without re-architecting our tools." — Agata Rother, Operations Systems and Insights Team Lead, Wayve

Agent Bricks Launch Partner

We're also a launch partner for Agent Bricks, Databricks' platform for building and governing enterprise AI agents grounded in your own data. New capabilities like Document Intelligence for unstructured data, Genie Agent Mode for multi-step reasoning, and Custom Agents give joint customers more ways to put AI to work on real business problems. The integration that has our customers most excited is our agent-to-agent connection with Genie Spaces. When Databricks surfaces a signal like a revenue anomaly, an inventory gap, or a support spike, your teams can act on it directly inside Retool without switching tools or rebuilding context.

“We can now move from explanation to action without losing momentum. Agent-to-agent communication between Databricks Genie Spaces and Retool means the right people are responding to the right signals sooner, on governed enterprise data, and influencing high-stakes decisions." — Anthony Perez, VP of Innovation and Growth Programs, MiQ

Retool sessions at Data + AI Summit 2026

If you're at Summit this week, we have three sessions worth catching:

Find us at booth #212

The team is on the ground all week at Moscone. Stop by Booth #212 to see live demos and learn about the exciting new way we’re making enterprise vibe coding safe for everyone.

Connie Wu
Connie Wu
VP of Global Partnerships, Retool
Connie oversees Retool’s global partner ecosystem, including partner strategy, operations and go-to-market activities. She has spent her career taking new technologies to market and driving transformation at companies like Asana, DocuSign and PayPal.
Copied