Tricentis, an Austin, Texas-based software testing company, announced on July 30, 2026 that it had acquired Tabnine, the Israeli AI coding platform originally founded as Codota in 2013.
The companies did not disclose financial terms. Israeli business outlet Calcalist estimated the price at tens of millions of dollars, well within the $55 million Tabnine had raised from investors.
Tricentis said it would integrate Tabnine's Enterprise Context Engine into its Agentic Quality Engineering Platform, where it will feed the company's AI testing and quality agents. According to Calcalist, existing Tabnine customers will continue to receive support for the products they use today. Neither company said what would happen to Tabnine's staff.
What Tabnine built
Tabnine was founded by chief executive Dror Weiss and chief technology officer Eran Yahav, a computer science professor. The company started out as Codota and in 2019 bought a Canadian startup called Tabnine, taking its name. Calcalist described it as one of the first commercial AI coding assistants, launched before GitHub Copilot.
The company later moved beyond code completion. Its Enterprise Context Engine builds a continuously updated knowledge graph of a company's systems, pulling entities, dependencies, relationships and architectural patterns out of code repositories, documentation, tickets, APIs and infrastructure metadata. The aim is to give AI agents an understanding of how a company's software fits together before they act on it.
Tabnine said organisations using the engine had reported up to twice the AI accuracy, up to 80% lower token consumption and up to 50% faster resolution of complex tasks. These are company-reported figures and were not independently verified.
“Quality engineering in the enterprise has never been a model problem. It has always been a context problem.”
Why Tricentis is buying
Tricentis sells tools for testing enterprise software and says it has more than 3,000 customers, including Experian, T-Mobile, Jaguar Land Rover, Allianz and Telstra. In its announcement the company argued that AI agents cannot reliably test or validate software they do not understand, and that missing context leads to wrong decisions and false confidence in release quality.
The Tabnine technology is meant to fill that gap. Tricentis said it would use it for modelling enterprise systems with a combined graph and vector approach, tracing dependencies so agents can see the downstream effect of a change, checking agent output against architecture and coding standards, and providing shared memory for multiple agents working together. The engine can run on-premises, in a private cloud or in fully air-gapped environments.
Tricentis chief executive Kevin Thompson said enterprise quality engineering had always been a context problem rather than a model problem, and that Tabnine's context layer belonged at the centre of how Tricentis delivers software quality. Dror Weiss said the engine was built so that agents understand the systems they operate in before they act, and that the Tricentis platform was what it had been designed for.
The founders and the funding
Tabnine raised $55 million in total, according to Calcalist and TechCrunch. Its last reported round was a $25 million Series B in November 2023 led by Telstra Ventures, with Atlassian Ventures joining as a new investor alongside existing backers Elaia, Headline, Hetz Ventures, Khosla Ventures and TPY Capital. Calcalist also lists OurCrowd among its investors.
At the time of the Series B, TechCrunch reported that Tabnine had about 60 employees and claimed more than one million users and 10,000 customers. Calcalist reported that the company later cut 15 of its roughly 80 employees to concentrate on enterprise AI products.
What changes for customers
The announcements focused on the context engine rather than Tabnine's standalone coding assistant. Analysis site Beri noted that neither company addressed the assistant's roadmap, pricing or deployment options, and that Tabnine's pricing page still listed monthly per-user plans at the time. The only public commitment was continued support for current products.
Tricentis is backed by Insight Partners and GTCR, which invested $1.33 billion in the company in November 2024 at a $4.5 billion valuation, according to Beri. The Tabnine deal added a coding-context layer to a testing platform at a time when software teams were handing more development and testing work to AI agents.
