Tricentis Acquires Israeli AI Coding Startup Tabnine for Tens of Millions of Dollars

Tricentis, a provider of software test automation platforms for large enterprises, announced on July 30, 2026 the acquisition of Israeli startup Tabnine, known for its AI coding assistant. According to Israeli outlet Calcalist, the deal is valued at tens of millions of dollars.

For years Tabnine was known mainly for code completion in developers' editors, competing with GitHub Copilot and other coding assistants. But Tricentis isn't buying that feature for its own sake. What the company wants is a technology Tabnine developed in parallel: the Enterprise Context Engine, an engine that builds a structured, continuously updated knowledge graph of an organization's systems.

What Tricentis actually bought

The Enterprise Context Engine extracts entities, dependencies, and architectural patterns from code repositories, documentation, service tickets, APIs, and infrastructure metadata. In practice, this produces a map of how a company's various system components connect to and depend on one another.

Tricentis wants to plug this engine into its Agentic Quality Engineering platform, a set of AI agents responsible for testing and quality assurance in large, complex enterprise environments. Rather than improving the language model itself, the company is betting on giving it better context about the system being tested.

Why context, not the model

Tricentis' argument comes down to a single problem: AI agents tasked with independently testing or validating software cannot do it well without understanding the full context of the system they operate in. Without that knowledge, autonomous agents can make bad decisions, introduce risk, and create a false sense of confidence about release quality.

Quality engineering in the enterprise has never been a model problem. It has always been a context problem - Kevin Thompson, CEO of Tricentis

We built the Enterprise Context Engine because AI in the enterprise is only valuable when it is reliable - Dror Weiss, founder and CEO of Tabnine

The numbers Tricentis is touting

The company cites results reported by customers testing the combined technologies: up to double the accuracy of AI responses, token usage cut by as much as 80 percent, and complex tasks resolved up to twice as fast. These are concrete, measurable claims Tricentis hopes will convince enterprise IT departments to adopt agentic software testing instead of classic, manually written test scripts.

Market context

The deal fits into a broader consolidation trend in AI-powered developer tooling. Major players in adjacent fields (testing, DevOps, quality management) are increasingly buying smaller AI startups not to acquire their consumer product, but to absorb a specific technology that strengthens their own agentic platform.

For Tabnine, this closes the chapter on independent growth. The company raised a total of $102 million from investors including Qualcomm Ventures, OurCrowd, Samsung NEXT Ventures, and Khosla Ventures since its first funding round in 2017. It most recently employed about 68 people.

For the Polish market, where more and more companies are deploying coding and testing agents in the daily work of IT teams, the deal signals that value in this industry increasingly lies not in the language model itself but in the data and context layer a vendor can deliver to the agent. That could shape which tools Polish development and QA teams choose when evaluating enterprise vendor offerings.