# Plato Raises $16 Million to Build Simulated Workplaces Where AI Labs Train Agents, and Checks Whether Models Game the Score

Times of AI Desk · 2026-10-28 · Research

[https://timesof.ai/2026/10/plato-16m-series-a-simulated-workplaces-agent-training](https://timesof.ai/2026/10/plato-16m-series-a-simulated-workplaces-agent-training)

> Plato, a San Francisco startup founded in 2025, said on October 28 it raised a $16 million Series A led by Nava Ventures, with Gradient Ventures, Transpose Platform and Microsoft AI CEO Mustafa Suleyman participating, bringing total funding to $19 million. It builds resettable simulations of jobs in customer support, finance, accounting, people operations and healthcare administration that AI labs use for reinforcement learning. The company says its evaluations check how a model reached a result, including whether it exploited weaknesses in the simulation. It named no lab customers.

In Plato's pitch, the simulated office matters less than the grading. The startup says a model that completes a purchase should not count as successful if it went over budget, and that its evaluations check whether models exploit weaknesses in a simulation to earn a high score. That is the behaviour AI researchers call reward hacking.

## The round

Plato said on Wednesday it raised a $16 million Series A led by Nava Ventures, with participation from Gradient Ventures, Transpose Platform and Mustafa Suleyman, whom the release identifies as Microsoft AI's chief executive. Nava founder Freddie Martignetti joins the board, and the round brings Plato's total funding to $19 million after a seed round led by Gradient Ventures. The company was founded in 2025 and is based in San Francisco.

## What it builds

Plato builds simulated versions of jobs in customer support, people operations, finance, accounting and healthcare administration, working with domain experts, and AI labs use them for reinforcement learning, according to the company. The simulations can be reset and reused, so a model learning to place an order does not spend real money and one practising an administrative task does not change real records, Plato says. It is developing longer sequences of work and interactions with simulated people.

"We build evaluations that reflect what the job actually requires, so developers can distinguish a high score from work done correctly," said co-founder and chief technology officer Pranav Putta. Chief executive Robert Farlow previously built an annotation platform at Yutori for training computer-use models, and Putta worked on training such models at MultiOn, the release says.

Plato says it will use the money to expand its library by paying professionals to recreate their jobs in simulated environments and by deploying AI agents on real tasks to collect more experience.

## Limits

- This account rests on Plato's release; we found no independent report we could read. Axios Pro's coverage is paywalled.
- Plato does not name the AI labs that use its environments, and gives no figures on how much training improves model performance.
- The release does not say whether Suleyman invested personally or through Microsoft; it lists him by his Microsoft AI title.

## Sources

- [PR Newswire: Plato Raises $16 Million to Build the Training Grounds for AI's Next Leap Into Real Work (October 28, 2026)](https://www.prnewswire.com/news-releases/plato-raises-16-million-to-build-the-training-grounds-for-ais-next-leap-into-real-work-302921832.html)