Analysis
Plato, a San Francisco startup that builds simulated workplaces for AI agents to practice real jobs, raised a $16 million Series A led by Nava Ventures, with Microsoft AI chief executive Mustafa Suleyman joining as an individual investor, the company announced Oct. 28.
What Plato does
Founded in 2025, Plato builds working simulations of real jobs and tasks that AI labs use as reinforcement-learning environments, according to PR Newswire. Co-founder and CEO Robert Farlow previously built an annotation platform at Yutori for training computer-use models. Co-founder and CTO Pranav Putta has a research background in computer vision, robotics and reinforcement learning, and worked on training computer-use models at MultiOn. Plato's approach is to pay real professionals to recreate their own jobs inside simulations -- covering customer support, people operations, finance, accounting and healthcare administration -- then deploy AI agents on those recreated tasks to generate training data. "Mastering real complex work comes from experience doing those tasks," Farlow said.
Plato funding history
Plato previously raised a seed round led by Gradient Ventures, Google's AI-focused venture fund; the amount wasn't disclosed. The new $16 million Series A brings Plato's total funding to $19 million. Nava Ventures founder Freddie Martignetti is joining Plato's board as part of the round. Martignetti described the company's role expansively: "we quickly learned that Plato is really a data lab."
“The new $16 million Series A brings Plato's total funding to $19 million.”
Who invested in Plato
Nava Ventures led the Series A. Gradient Ventures, which led the earlier seed, returned to participate again, alongside new investor Transpose Platform. Microsoft AI chief executive Mustafa Suleyman joined as an individual investor -- a notable personal bet from a sitting Big Tech AI leader on a company building the training infrastructure other labs, potentially including Microsoft's own AI division, might rely on.
AI training-data market and competitors
Plato sits inside the reinforcement-learning-environments market, per Fortune.
- Market size -- ResearchAndMarkets estimates roughly $11.6 billion in 2025 growing to more than $90 billion by 2034, though that figure spans the broader category of RL tools and environments, not environment-building startups alone.
- Deeptune -- a comparable startup building high-fidelity RL environments, raised a larger $43 million Series A led by Andreessen Horowitz earlier in 2026.
Plato's approach, paying professionals to recreate their actual jobs rather than relying on synthetic or scraped data, differentiates it from environment-builders that generate simulations algorithmically, part of the fast-growing AI infrastructure landscape we track in our funding tracker.
What to watch next
The open question is whether Plato's human-recreated job simulations produce measurably better-trained agents than cheaper synthetic alternatives -- if not, the labor-intensive data-collection model becomes a cost disadvantage rather than a moat. Suleyman's personal investment is a signal worth watching for whether Microsoft's own AI teams end up as a Plato customer, which would validate the model faster than any funding round could. Plato's $19 million in total funding is still modest next to Deeptune's standalone $43 million Series A, so the near-term task is proving its human-recreated environments are worth the premium they cost to build.