Open-source Python framework (and managed enterprise platform) for orchestrating multi-agent workflows: 'crews' of role-based AI agents with shared goals, tools and memory.

Business automation will be built from teams of collaborating AI agents — an open-source Python framework anyone can adopt — not from one model doing everything.

$18M announced Oct 22, 2024: inception/seed led by boldstart Ventures and Series A led by Insight Partners, at a ~$100M valuation, with angels including Andrew Ng, Dharmesh Shah and Sam Altman's family fund Alt Capital.

João Moura directed AI engineering at Clearbit until HubSpot acquired it, and had founded startups before (smart pet-collar maker Urdog). In October 2023 he began CrewAI as a side project, open-sourcing the role-based agent framework in November 2023 and launching the company in January 2024, betting that collaborating specialized agents — not single-model automations — would replace brittle RPA.

CrewAI repeatedly hit GitHub Trending: Python #1 on 2025-09-02 and 2025-09-04, Python #5 on 2026-04-23, passing 44k stars by Feb 2026 and 57.9k by Sep 2026. It signed 150 enterprise customers in its first year and claimed 100,000 multi-agent executions per day by October 2024, when it raised $18M (boldstart inception, Insight Series A) at ~$100M valuation and launched CrewAI Enterprise, later the commercial AMP Suite, with an IBM watson.ai partnership. An Insight Partners profile counted 450M agents executed per month by Dec 2025; customers include PwC, IBM, Capgemini and NVIDIA.

Still scaling as of 2026-09-01: The Agent Times reports nearly half the Fortune 500 deploys agents through the framework, 100k+ certified developers on its courses, and the company is selling a commercial control plane (AMP Suite) around the open-source core.

Open source made distribution free: developers adopted the framework on GitHub before there was a sales team.

The bet landed on a structural gap — multi-agent orchestration — as single-agent demos kept hitting real-world limits.

Famous angels and an IBM partnership turned open-source popularity into enterprise trust quickly.

The framework abstracted models, so customers could switch LLM vendors without rewriting automation, matching how buyers actually evaluate AI.

Bet on the layer beneath the app: open-sourcing the orchestration framework made every developer's agent a distribution channel, and GitHub rank became enterprise credibility.

As of 2026-09-01 CrewAI is scaling from San Francisco and Brazil with 57.9k GitHub stars, 450M agents executed per month, and a user base that includes PwC, IBM, Capgemini and NVIDIA. It sells CrewAI AMP Suite — managed deployment, observability and governance — as its commercial layer, runs a certification program with 100k+ developers, and pushes the open-source framework with event-driven Flows alongside Crews. Competition includes LangGraph and AutoGen, but CrewAI remains the highest-profile open-source multi-agent standard.

FOLLOW THE EVIDENCE

The sources

  1. CrewAI uses third-party models to automate business tasks TechCrunch · techcrunch.com ·
  2. CrewAI Hits 44K Stars and 450 Million Agents Per Month. The Orchestration Layer Is Here. The Agent Times · theagenttimes.com ·
  3. crewAIInc/crewAI — 57.9k Stars, Global Rank #387 star-history.com · star-history.com ·
  4. crewAIInc/crewAI repository GitHub · github.com