Mercor acquires Deeptune, which builds reinforcement learning environments for AI agents, three months after CEO Brendan Foody backed Deeptune's $43M Series A
FortuneLily Mae Lazarus
Context & Ripple Effects
Mercor has moved from an AI-vetted jobs marketplace toward supplying domain experts for model training, while Deeptune raised a $43M Series A to create high-fidelity reinforcement-learning environments modeled on professional workflows. The acquisition links Mercor’s access to expert work with a platform designed to turn such work into agent-training settings.
The deal follows CEO Brendan Foody’s backing of Deeptune’s Series A only three months earlier, indicating that Mercor had early visibility into the company before bringing it inside the business.
First-order effects
Mercor gains control of Deeptune’s reinforcement-learning environment technology and team, adding a training-environment capability alongside its expert-sourcing operation.
Deeptune’s standalone path changes: its products and development priorities can now be directed toward Mercor’s model-training and agent-workflow efforts.
Second-order effects
Mercor can more closely connect domain experts with simulated professional tasks, potentially making its training-data offering more differentiated than a service focused only on recruiting or labeling.
Rivals in AI training services, including companies competing with Mercor for experts, face added pressure to show they can provide not just talent but the environments in which agents can be trained and evaluated.
Third-order effects
If similar combinations continue, the AI-training market could consolidate around vertically integrated providers that control expert supply, workflow simulation, and reinforcement-learning feedback rather than selling those components separately.
The key uncertainty is whether high-fidelity workflow environments become a durable bottleneck for agent development; if they do, ownership of proprietary environments and expert-generated task design may become strategically important.
The trend: This is part of a shift from sourcing human input for models toward building integrated infrastructure for training and testing AI agents on professional work.
Excited to share that @deeptuneai is joining @mercor_ai. When we started Deeptune, we believed RL environments would become the defining bottleneck for AI. Over the past two years, we've seen this play out even faster than we expected. Mercor has built the world's leading [image]
I'm excited to announce that @mercor_ai is acquiring @deeptuneai. While Mercor is the largest RL Environments company, Deeptune is the top vendor to Frontier Labs for the apps that make those environments work. The barrier to automating everything you do on your laptop with
Excited for Deeptune and Mercor to join forces. Data is AI's major bottlenecks, and both teams are trailblazers in the space: Mercor for its extensive expert network and Deeptune for their industry leading RL environments. They'll go faster and further together Working with Tim […
I'm so proud of @timlupo and the entire @deeptuneai - he's a brilliant and thoughtful researcher we were lucky to lead the seed round of a few years ago and what they've accomplished (especially in this last 2 years of breakout) has helped AI leap forward https://x.com/...
Mercor is acquiring @deeptuneai. Training AI to do real professional work requires covering the full distribution of apps, worlds, and tasks across the economy. We have built the largest expert network and have been a leader in creating worlds and tasks. Deeptune built the apps,
I'm so proud of Tim and his team at DeepTune - we led their seed round years ago and they grinded through the darkness to arrive at something VERY important a couple years ago. …