Freeplay is a Boulder, Colorado-based company building a unified platform for teams developing AI products powered by large language models. Its founders, Ian Cairns and Eric Ringger, previously led product and engineering for Twitter's developer platform and saw firsthand how hard it is to manage software when behavior is probabilistic. Freeplay's premise is that building good LLM features is a team sport spanning product managers, engineers, and subject-matter experts, and that these groups need a common workflow rather than scattered scripts and spreadsheets.
The platform brings together several capabilities that are often handled by separate tools. Prompt and model management lets teams version prompts, swap models, and collaborate without redeploying code. The evaluation system combines automated 'auto-evaluators' — LLM-as-a-judge scorers configured for a team's specific quality criteria — with human review and labeling, so subjective quality can be measured at scale. Experimentation tooling allows teams to compare prompts, models, and configurations on curated datasets before shipping. Once in production, Freeplay provides observability to monitor real traffic, capture interesting or failing cases, and feed them back into test sets.
This closed loop — prototype, test, ship, monitor, and improve — is Freeplay's core value proposition. By making it easy to turn production failures into evaluation cases and to involve non-engineers in defining and labeling quality, the platform helps teams iterate faster and trust what they release to customers.
Freeplay raised an initial seed round in late 2023 co-led by Conviction and Matchstick Ventures, and a further round of roughly $5.6M in 2025 led by Renegade Partners, with participation from Conviction, Matchstick, Next Frontier, and PWV, alongside angels from GitHub, Dropbox, Atlassian, and Scale AI. The company targets product-led organizations that want to move beyond prompt tinkering to a disciplined, measurable AI development practice.