Composio Raises $29M to Solve AI's Learning Problem: Building Skills That Actually Improve Over Time
SAN FRANCISCO, July 22, 2025 /PRNewswire/ -- Composio today announces $25 million in Series A funding led by Lightspeed Venture Partners to tackle the fundamental limitation preventing AI agents from transforming enterprise workflows: they don't learn from experience.
Several industry leaders also participated in the round, including Guillermo Rauch (Vercel CEO), Dharmesh Shah (HubSpot CTO & Founder), Gokul Rajaram, and Soham Mazumdar (Rubrik Co-founder), alongside institutional investors SV Angel, Blitzscaling Ventures, Operator Partners, and Agent Fund by Yohei Nakajima. Existing investors Elevation Capital and Together Fund also participated, bringing the total funding to $29 million.
The Real Reason AI Agents Aren't Transforming Work Yet
Despite the hype, Fortune 500 companies aren't using AI agents to transform their workflows, not because management is resistant to change, but because current AI lacks a fundamental capability: the ability to improve through practice.
"You can spend hundreds of hours building LLM tools, tweaking prompts, and refining instructions, but you hit a wall," says Soham Ganatra, CEO of Composio. "These models don't get better at their jobs the way a human employee would. They can't build context, learn from mistakes, or develop the subtle understanding that makes human workers invaluable. We're solving this at the infrastructure level."
Composio's approach centers on creating a shared learning layer for AI agents. When an agent on the platform learns how to handle a Salesforce integration edge case, optimize a GitHub workflow, or scaffold a database using Supabase MCP, that knowledge becomes available to every other agent—creating a network effect where the entire ecosystem improves together.
Building the Missing Infrastructure
The company started building Composio two years ago, before AI agents became the current hype, recognizing that making agents truly useful would require solving fundamental infrastructure problems. The team tackled complex challenges including multi-agent coordination, authentication across enterprise systems, and building scalable infrastructure that now processes millions of requests daily.
Most importantly, they developed a reinforcement learning layer that enables AI to build intuition from experience—something missing from conventional approaches.
"The challenge isn't making AI smarter in isolation," explains Ganatra. "It's giving AI the ability to accumulate practical knowledge the way humans do—but at the scale and speed only software can achieve."
Early Traction Points to Market Demand
The platform has attracted over 100,000 developers, with adoption accelerating among AI-first companies. Top companies from the latest YC batches like April, OpenNote, Airweave, Den, and Dash have chosen Composio. Popular AI launches like Context and Altera have also built on the platform. The company serves over 200 startups and enterprises including Glean, already generating seven-figure revenue.
A typical use case: instead of spending months building authentication, handling edge cases, and debugging integrations, developers can tap into Composio's existing skills to ship their agents within days. More importantly, their agents don't start from scratch, they immediately benefit from what thousands of other agents have already learned.
"What excites us about Composio is that they're not just solving today's integration problems," said Raviraj Jain from Lightspeed Venture Partners. "They're building the foundation for AI agents to become genuinely useful by learning from experience at scale. This is the missing piece between impressive demos and transformative deployments."
The Future: Creating AI That Actually Learns
With this new funding, Composio will accelerate development of its learning infrastructure. Unlike conventional AI agents that remain static despite repeated use, Composio's technology enables systems to continuously improve by accumulating experience.
By building a foundation that captures practical knowledge and develops contextual understanding, Composio is transforming AI from rigid tools into adaptive partners capable of genuine expertise. The platform seamlessly integrates with all major AI frameworks, including MCP, LangChain, Vercel AI SDK, and OpenAI Agents, allowing developers to implement these learning capabilities regardless of their preferred technology stack.
About Composio
Composio provides the learning infrastructure that enables AI agents to improve through experience. By creating a shared skill layer that captures and distributes practical knowledge across the entire AI ecosystem, Composio transforms static AI tools into systems that develop real expertise over time. For more information, visit https://composio.dev/
View original content:https://www.prnewswire.com/news-releases/composio-raises-29m-to-solve-ais-learning-problem-building-skills-that-actually-improve-over-time-302510684.html
SOURCE Composio