How HawkenAI built AG2's MCP marketplace and agent platform, hosting more than 100,000 public MCP servers in 30 days with infrastructure engineered to scale without limits.
AG2 is the open-source framework for building, orchestrating, and scaling multi-agent AI systems, created by the team behind AutoGen. Open source at the core and enterprise-ready at scale, AG2 is trusted by enterprise teams and leading research institutions, and its framework is used by millions of developers.
As the agent ecosystem exploded, AG2 set out to give developers something the market did not yet have: a single place to discover, host, and orchestrate the full universe of agent tools. The Model Context Protocol became the connective tissue of the agent era, but thousands of MCP servers were scattered across the internet with no unified home.
AG2 wanted to close that gap with an MCP marketplace and agent platform that hosted the entire public ecosystem, with the infrastructure to run it reliably and scale without limits. Hosting every public MCP server means solving for massive scale, reliability, and performance from day one, and doing it fast enough to matter in a market moving this quickly.
One unified place to discover, host, and orchestrate MCP servers instead of a fragmented public ecosystem.
Infrastructure to host the full public MCP catalog, not just a handful of popular servers.
Enterprise-grade reliability and performance from day one, with room to scale without compromise.
Speed to market fast enough to matter as the agent ecosystem accelerated.
HawkenAI designed and built a complete MCP marketplace and AI agent platform for AG2, including the underlying infrastructure and the scalability to support the entire public ecosystem.
HawkenAI built the full marketplace experience and agent platform AG2 needed to unify discovery, hosting, and orchestration in one product.
The platform hosts every public MCP server, giving developers one place to find and run any agent tool available in the ecosystem.
HawkenAI owned the cloud infrastructure engineered to host and serve MCP servers at massive scale with the reliability enterprise teams expect.
HawkenAI created and launched the marketplace in a single month, standing up an ecosystem of that size and complexity at rare engineering velocity.
The platform is model-agnostic by design, matching AG2's philosophy of orchestrating the best agents and tools regardless of origin. HawkenAI built it to use multiple frontier models, choosing the right model for each job across the agent workflows the platform powers.
The platform orchestrates the best agents and tools regardless of origin, giving AG2 and its developers flexibility across workflows.
Frontier Claude models power deep reasoning tasks across agent workflows where accuracy and nuance matter most.
Models from OpenAI extend the platform's reach, letting AG2 match the strongest available capability to each task.
Multi-model architecture delivers flexibility, resilience, and access to the strongest capabilities available for any given agent workflow.
The platform has been live in production since August 2025. In 30 days, HawkenAI delivered a marketplace hosting the full public MCP ecosystem of more than 100,000 servers, on infrastructure built to scale.
That foundation feeds directly into the impact AG2's platform delivers for the organizations that build on it. The marketplace and agent infrastructure HawkenAI built give developers the reach and reliability that make outcomes like these possible.
AG2 reports up to 70 percent faster workflows and up to 5x productivity gains across its platform, with full decision auditability for enterprise teams.
The marketplace has been live in production since August 2025, putting the entire public MCP ecosystem within reach and running it at scale.
AG2 needed the entire public MCP ecosystem in one place, with infrastructure that could scale without compromise, and they needed it fast.
“We needed to bring the entire public MCP ecosystem into one place, with infrastructure that could scale without compromise, and we needed it fast. HawkenAI delivered a marketplace hosting more than 100,000 servers in 30 days. The speed and the engineering quality were both exceptional. It gave our developers the reach and reliability that a platform at our scale demands.”
HawkenAI is an engineering and applied AI consultancy that builds production AI platforms for growth-stage companies and enterprises. HawkenAI partners with teams to design, build, and ship AI systems that create measurable business value, with a focus on real outcomes, serious scale, and fast execution.
From core intelligence layers to customer-facing AI applications, HawkenAI ships products that move from prototype to operational use.
The focus is measurable leverage: faster decisions, lower operational drag, and AI that fits the business instead of adding complexity around it.
If you’re ready to accelerate your roadmap, outpace competitors, and get to market faster than ever, let’s talk.
HawkenAI customer story about building AG2's MCP marketplace and agent platform, hosting more than 100,000 public MCP servers in 30 days with infrastructure engineered to scale.
AG2 is the open-source framework for building, orchestrating, and scaling multi-agent AI systems, created by the team behind AutoGen. Open source at the core and enterprise-ready at scale, AG2 is trusted by enterprise teams and leading research institutions, and its framework is used by millions of developers. As the agent ecosystem exploded, AG2 set out to give developers something the market did not yet have: a single place to discover, host, and orchestrate the full universe of agent tools.
HawkenAI built it.
The Model Context Protocol turned into the connective tissue of the agent era. Thousands of MCP servers appeared, each exposing tools, data, and capabilities that agents can use. But they were scattered across the internet with no unified home, no reliable hosting, and no consistent way to bring them into an agent system.
AG2 wanted to close that gap with an MCP marketplace and agent platform that did not just index a handful of popular servers, but hosted the entire public ecosystem, with the infrastructure to run it reliably and scale without limits. That is a serious engineering problem. Hosting every public MCP server means solving for massive scale, reliability, and performance from day one, and doing it fast enough to matter in a market moving this quickly.
HawkenAI designed and built a complete MCP marketplace and AI agent platform for AG2, including the underlying infrastructure and the scalability to support the entire public ecosystem.
The platform hosts every public MCP server, more than 100,000 in total, giving developers one place to find and run any agent tool available. HawkenAI owned the full build: the marketplace experience, the agent platform, and the cloud infrastructure engineered to host and serve MCP servers at that scale with the reliability enterprise teams expect.
The pace set this engagement apart. HawkenAI created and launched the marketplace, with more than 100,000 MCP servers, in 30 days. Standing up an ecosystem of that size and complexity in a single month is a rare feat of engineering execution.
The platform is model-agnostic by design, which matches AG2's philosophy of orchestrating the best agents and tools regardless of origin. HawkenAI built it to use multiple frontier models, including Claude from Anthropic and models from OpenAI, choosing the right model for each job across the agent workflows the platform powers.
This gives AG2 and its developers flexibility, resilience, and access to the strongest capabilities available for any given task.
The platform has been live in production since August 2025. In 30 days, HawkenAI delivered a marketplace hosting the full public MCP ecosystem of more than 100,000 servers, on infrastructure built to scale.
That foundation feeds directly into the impact AG2's platform delivers for the organizations that build on it. AG2 reports up to 70 percent faster workflows, up to 5x productivity gains, and full decision auditability across its platform. The marketplace and agent infrastructure HawkenAI built give developers the reach and reliability that make outcomes like these possible, by putting the entire agent toolset within reach and running it at scale.
“We needed to bring the entire public MCP ecosystem into one place, with infrastructure that could scale without compromise, and we needed it fast. HawkenAI delivered a marketplace hosting more than 100,000 servers in 30 days. The speed and the engineering quality were both exceptional. It gave our developers the reach and reliability that a platform at our scale demands.”
HawkenAI is an engineering and applied AI consultancy that builds production AI platforms for growth-stage companies and enterprises. HawkenAI partners with teams to design, build, and ship AI systems that create measurable business value, with a focus on real outcomes, serious scale, and fast execution. Learn more at hawken.ai.
