A New Free Tool Called Proliferate Wants Teams to Work With AI Coding Assistants Together
A new company called Proliferate has launched a free, open-source tool that helps teams of programmers work alongside AI coding assistants. The San Francisco startup has financial backing from True Ventures, Pear VC, and more than 50 Y Combinator founders. The company was founded in 2025 by Pablo Hansen and is currently hiring a founding product engineer through a Y Combinator job listing posted August 26, 2026 (Y Combinator job listing).
Hansen previously joined Onyx (YC W24) as its first engineer, where he worked on enterprise AI search. Proliferate operates under the tagline "Do your life's work with agents." The team works in person in SoMa, San Francisco.
The founding product engineer role is full-time in San Francisco, paying $150K-$300K with 0.30%-1.50% equity. Visa sponsorship is available, and the role is open to new graduates. The position places the hire among the first engineers working alongside Hansen and the early team.
On the product side, Proliferate works with several AI coding assistants, including Claude Code, Codex, and Gemini, using the developer's existing subscriptions rather than requiring separate model licensing (Proliferate website). The software is licensed under AGPL-3.0, and the desktop app is available for Mac download at no cost. Proliferate states the desktop app will remain free and open source.
For teams, Proliferate is developing a private beta that includes shared workspaces, review controls, and the ability to coordinate multiple AI assistants at once. The platform offers each task its own isolated copy of the code and its own temporary workspace for running tests, and it integrates with GitHub, Slack, and Linear. Companies can also host the software on their own servers, inside their own private cloud environment.
Some of these design choices are worth a closer look. Giving each AI assistant its own separate copy of the code means that multiple assistants running at the same time cannot step on each other's changes. The temporary workspaces let each assistant run builds and tests without affecting anyone else's work. This is the same idea behind how automated testing systems have worked for years, applied here to AI assistants instead of testing pipelines.
The AGPL-3.0 license carries a specific implication worth flagging. It requires that any organization running a modified version of the software over a network must publish its modifications. This creates a tension: companies that want to customize Proliferate for internal use and share it with employees over a network would, strictly speaking, have to release their changes publicly. Whether that becomes a friction point or an acceptable trade-off will depend on how aggressively Proliferate enforces the license and whether it offers a commercial license as an alternative. The facts available do not indicate a commercial license tier, but the private beta for teams suggests a managed or hosted offering is in development.
The model-agnostic approach also carries trade-offs. By supporting Claude Code, Codex, and Gemini on users' existing subscriptions, Proliferate positions itself as a coordination and collaboration layer rather than an AI provider. This avoids the cost pressure of reselling access to AI models, but it also means the product's value must come entirely from its workflow, isolation, and team coordination features rather than from any proprietary AI capability.
The broader context is that the AI-assisted coding tool space has become crowded, with tools ranging from GitHub Copilot's editor integrations to standalone agent runners. What sets Proliferate apart is the combination of open-source licensing, the ability to host it on your own servers, and the explicit framing of AI assistant collaboration as a team activity rather than a solo experience. The private beta features, shared workspaces and review controls in particular, address the organizational question of how teams review and approve AI-generated code before it reaches their main codebase.
The backing from True Ventures and Pear VC, plus participation from 50+ YC founders, gives Proliferate funding and network access typical of well-resourced early-stage companies in the current AI tooling cycle. Hansen's prior experience at Onyx, building enterprise AI search, provides relevant background in shipping AI-powered tools to organizations with procurement and security requirements.
The hiring signal is straightforward: a founding engineer at an early-stage company with strong investor backing, open-source distribution, and an in-person team in San Francisco. The compensation range of $150K-$300K with meaningful equity is consistent with early-stage YC company norms. The visa sponsorship availability broadens the candidate pool beyond domestic applicants, and the openness to new graduates suggests the company prioritizes capability over seniority for this particular role.
For anyone tracking the AI tooling ecosystem, Proliferate is one to watch if the idea of open-source, self-hostable, multi-agent team workflows catches on. The product is available now for Mac, with team features in private beta.
