
Poolside envisions a future where AI fundamentally transforms software development, making coding accessible to billions of people worldwide and unlocking new scientific and technological breakthroughs. We are committed to building advanced AI systems that learn and create code autonomously, moving humanity closer to artificial general intelligence capable of solving complex, real-world problems.
Harnessing proprietary large language models and innovative reinforcement learning techniques, Poolside drives innovation in AI-assisted software creation, empowering developers with scalable, secure tools integrated into leading cloud environments. Our work is not just about writing code; it’s about advancing the frontier of intelligent machines to accelerate progress across science, health, and education.
With a mission centered on making software development more efficient and inclusive, Poolside is pioneering an AI future where technology adapts and evolves through continuous learning, amplifying human creativity and addressing challenges at both individual and enterprise scales worldwide.
Our Review
When GitHub's former CTO Jason Warner teams up with source{d}'s founder Eiso Kant to build an AI research lab, we pay attention. Poolside isn't just another coding assistant — it's an ambitious bet on fundamentally changing who can build software and how they do it.
The company's mission hits different than most AI startups. Sure, they want to help the billions of people who can't code today. But their real moonshot? Building AGI that can tackle complex software systems and eventually solve major scientific challenges. It's bold, maybe even audacious.
What Makes Them Different
Here's where Poolside gets interesting: their AI doesn't just regurgitate code patterns. It actually learns through trial and error, using something called Reinforcement Learning from Code Execution Feedback (RLCEF). Think of it as an AI that writes code, runs it, fails, learns from the mistakes, and gets better — just like human developers do.
They've trained their models on over 500,000 open-source codebases, which is impressive scale. But the real kicker is that their AI can code autonomously and continuously improve itself. That's not your typical autocomplete tool.
The Enterprise Play
While consumer coding tools grab headlines, Poolside is laser-focused on enterprise customers. They're offering private, secure AI coding assistance that companies can deploy in their own environments. No data leaves your infrastructure, no competitive secrets get leaked.
Their partnership with AWS through Amazon Bedrock shows they're serious about enterprise-grade deployment. The usage-based pricing model makes sense too — you pay as you scale, which aligns perfectly with how enterprises actually adopt new tools.
Following the Money
$600 million in funding tells a story. That September 2024 $500 million Series B at a $2 billion valuation caught our attention — especially with Bain Capital Ventures leading and heavy hitters like Nvidia and eBay participating.
They're putting that money to work with serious infrastructure: 10,000+ Nvidia GPUs for training. That's not playing around — that's building for the long haul. When a company invests that heavily in compute power, they're signaling confidence in their technical approach and market opportunity.
Generative AI coding platform
AI coding assistant
API integration with cloud providers like AWS
Proprietary large language models
Reinforcement Learning from Code Execution Feedback (RLCEF)
Secure, private AI deployment
Performance and security testing support






