Mechanize co-founder Tamay Besiroglu has joined Google DeepMind, while more than a dozen former employees have moved to Google. Final financial terms remain undisclosed.
Google has completed a talent and technology deal with AI startup Mechanize, bringing co-founder and former CEO Tamay Besiroglu into Google DeepMind as a research scientist.
More than a dozen former Mechanize employees have also moved to Google, with many joining teams working on AI model “midtraining”, a stage focused on improving specific capabilities after a model’s initial training.
The deal follows earlier reports that Google was in talks for a transaction valued at more than $1.5 billion. However, the final financial terms have not been disclosed, so the reported $1.5 billion figure should not be treated as the confirmed price paid by Google.
Mechanize itself has not been absorbed as a conventional corporate acquisition. The arrangement has been described as a combination of technology licensing and hiring of key personnel, while the startup continues to operate separately.
Tamay Besiroglu Joins Google DeepMind
Besiroglu is now listed as a research scientist at Google DeepMind.
More than 12 other former Mechanize employees have also joined Google. Their work reportedly includes midtraining and related efforts to improve AI models’ ability to perform specialised tasks.
The move gives Google access to a team that has focused on building environments and evaluation systems for training AI models to handle complex, multi-step software engineering tasks.
Mechanize had positioned its technology around the development of AI systems capable of carrying out software engineering work with less continuous human supervision.
What Mechanize Was Building
Mechanize was founded in 2025 by Tamay Besiroglu, Matthew Barnett, and Ege Erdil.
Rather than developing another general-purpose chatbot, the startup focused on creating realistic environments where AI agents could perform and be evaluated on complex tasks.
These environments allow AI systems to work through software engineering problems, receive feedback, and repeat tasks in controlled settings.
The company has argued that realistic training environments and high-quality feedback are important limitations in developing more capable AI agents.
That technology is particularly relevant as AI companies work to move beyond systems that simply respond to prompts toward models capable of completing longer, more complicated workflows.
Google Paid More Than Mechanize’s Last Valuation?
Mechanize raised $9.1 million in funding earlier in 2026 at a reported $500 million valuation.
Reports in August said Google was considering a deal worth more than $1.5 billion for the startup’s technology and team.
The completed transaction, however, has not disclosed its final financial terms.
That distinction matters because the $1.5 billion figure refers to the reported value of negotiations rather than a confirmed purchase price for Mechanize.
Why AI Talent Is Becoming a Strategic Asset
The Mechanize deal reflects a growing trend in the technology industry in which major companies secure specialised AI researchers and engineers through talent-focused transactions rather than traditional acquisitions.
Google has used similar structures in deals involving AI companies including Character.AI and Windsurf.
For large AI labs, access to experienced researchers can be as important as acquiring a company’s products.
The Mechanize team brings expertise in training environments, evaluations, and the development of AI systems designed for complex software engineering tasks.
The Push Toward Autonomous AI Workers
Mechanize’s broader ambition was to automate software engineering and eventually expand AI-driven automation into other areas of knowledge work.
Its approach focused on AI agents that could operate for longer periods, complete multiple steps, and adapt to changing requirements rather than simply generate a single response.
The company’s founders have argued that large-scale automation of labour could create new economic opportunities and services while substantially changing how work is organised.
Besiroglu has previously argued that automating labour could lead to greater economic abundance and higher living standards.
For Google, bringing Mechanize’s talent into DeepMind adds another layer to its efforts to improve AI systems that can perform increasingly complex tasks.
Mechanize Remains a Separate Company
Despite reports describing the transaction as a $1.5 billion-plus deal, Mechanize has not simply disappeared into Google.
The reported structure combines technology access with the hiring of key employees, while the company remains a separate entity. Former Mechanize chief of staff Guive Assadi is listed as its CEO.
That makes the transaction different from a conventional acquisition in which Google would take ownership of the entire company.
Instead, Google has secured access to specialised technology and a significant portion of the startup’s talent while Mechanize continues independently.
What the Deal Means for Google’s AI Strategy
The Mechanize transaction highlights how the focus of AI development is expanding beyond larger models and more computing power.
Training environments, evaluation systems and techniques for improving specialised capabilities are becoming increasingly important as companies attempt to build AI agents that can handle real-world work.
For Google DeepMind, the addition of Besiroglu and more than a dozen former Mechanize employees strengthens its capabilities in this area.
The final financial terms of the deal remain undisclosed, but the reported $1.5 billion-plus negotiations underline the value that major AI companies are placing on specialised talent and technology as the race to build more capable AI systems accelerates.
Read more news and follow us on Instagram
Tamay Besiroglu, Co-Founder of Mechanize (Credit: MIT Future Tech)
Source: Business Connect
