Fireworks AI raises $250M at $4B valuation with Polish co-founder on board
Backed by the minds behind PyTorch, including Poland’s Paweł Garbacki, Fireworks AI secures $250M to reshape enterprise-grade AI infrastructure
In a standout moment for Europe’s tech diaspora, Fireworks AI, the California-based AI infrastructure company founded by engineers behind PyTorch, has raised $250 million in Series C funding at a $4 billion valuation. The round was co-led by Lightspeed Venture Partners, Index Ventures, and Evantic, with participation from Sequoia Capital. But what makes this milestone especially significant for Poland is the presence of Paweł Garbacki, a Warsaw-born engineer and one of Fireworks’ founding team members, marking one of the most notable Polish contributions to a soon-to-be AI unicorn in Silicon Valley.
This latest investment brings Fireworks’ total funding to $327 million and positions the company as a serious contender in the battle for AI infrastructure dominance.
Subscribe to Startup Digest to stay ahead with the latest news, investments, and must-attend events.
Enterprise-First, Developer-Centric
Fireworks is built around a powerful insight: enterprises need more than generic models; they need domain-specific AI that integrates with proprietary data and workflows. By delivering sub-second inference, developer-friendly tooling, and open-source model support, Fireworks allows companies like Uber, GitLab, and Shopify to own, train, and scale their own models without vendor lock-in.
Its technical foundation is formidable. Fireworks supports over 100 top-performing models across text, image, audio, and multimodal tasks and powers over 10 trillion tokens daily. Their serverless infrastructure, custom CUDA kernels, and model sharding pipeline outperform even hyperscalers like AWS and Google Cloud. The result? 12x faster inference than vLLM and 40x faster than GPT-4 benchmarks, at a fraction of the cost.
From Warsaw to the World’s AI Backbone
Paweł Garbacki, now based in the U.S., plays a key role in shaping Fireworks’ technical roadmap. A respected figure in the global PyTorch community, Garbacki’s journey from Warsaw’s academic and startup circles to co-building a $4B global AI infrastructure company illustrates the international reach and impact of Polish engineering talent. While many Polish startups remain local or EU-focused, Fireworks offers a glimpse of what Polish founders can achieve at the frontier of global AI.
A Platform for Builders, Not Black Boxes
Unlike closed platforms that lock users into proprietary APIs, Fireworks champions open innovation. Its product-model co-design system lets enterprises fine-tune performance based on real-world interactions. This feedback loop, combined with reinforcement learning and robust observability, ensures that companies don’t just deploy AI, they evolve it continuously.
Fireworks’ stack is especially attractive to AI-native startups that want full control over performance, data security, and latency. Partners like NVIDIA, Google Cloud, and MongoDB are already integrating with the platform, adding to its rapid adoption curve.
A Unicorn in Sight, and a Win for Polish Tech Diaspora
With this round, Fireworks is on a fast track to becoming one of the most important AI infrastructure providers in the world. For Poland, the success of Paweł Garbacki highlights how deep-tech talent cultivated in CEE can drive global category leaders, especially in foundational technologies like PyTorch and AI model optimization.
As European ecosystems seek to move beyond software applications toward core infrastructure innovation, Fireworks stands as a case study in global ambition backed by Polish roots, and a reminder that the next unicorns may not be built in Poland, but by Poles.
