The acquisition will enable enterprises to build agents that can leverage real-time data to make better decisions while acting autonomously, the company said.
To perform tasks as well as a human, AI agents need context — something that has been lacking in many systems to date, according to Databricks. It hopes to fill that gap with a new acquisition, that of San Francisco-based machine learning startup Tecton.
Databricks intends to use Tecton’s technology in its Agent Bricks offering to quickly turn data from its data lakehouse into context for AI agents to use, especially real-time data generated by event-driven data architectures.
“Without a clear system, preparing this data can be slow, repetitive, and prone to errors, making it tough for enterprises to move AI agents into production and slowing down innovation,” Databricks wrote in a blog post.
In contrast, it said, Tecton centralizes and automates the creation, sharing and serving of fresh and relevant contextual data to both traditional machine learning and AI agent systems in a fast, cost-efficient manner.
Analysts see enterprises benefiting from Tecton’s integration into Databricks.
“Enterprises can benefit from a streamlined workflow where data is transformed into production-ready features and immediately consumed by AI agents, accelerating the deployment of personalized, real-time applications and eliminating complex data engineering between separate systems,” said Charlie Dai, VP and principal analyst at Forrester.
IDC analyst and associate vice president Sharath Srinivasamurthy said that event-driven automation and analyzing data in motion is a critical enabler for enterprises desiring to scale their agentic systems. “When contextual, event-driven customer data and external signals are captured and analyzed in real time, the efficacy and responsiveness of AI agents significantly improves,” Srinivasamurthy said. Since event-driven agentic automation enables instantaneous decision-making based on the most current information, it not only improves responsiveness for time-sensitive workloads but also reduces compute waste, bringing down cost, he said.
“Agents can proactively respond to events such as errors, updates, or user requests, fostering real-time problem-solving and adaptive behavior. Examples where this is getting traction includes fraud detection, risk management, emergency response, predictive maintenance, dynamic pricing, traffic management, IT operations, and customer experience,” Srinivasamurthy added.
Tecton is Databricks’s fourth acquisition this year, and its second in the machine learning space following its April acquisition of feature engnineering specialist Fennel.
The Tecton deal comes just days after Databricks finalized a new investment round that, it said, will help it expand Agent Bricks to accelerate its AI strategy and push its new database offering, Lakebase. The investment values Databricks at over $100 billion, it said — although it did not say how much new capital would be injected.
