Stack AI Raises $16 Million to Build an AI Agent for Every Job

More than 82% of global companies are now using or evaluating artificial intelligence in their operations, and Stack AI is positioning itself to meet that demand not by building new language models, but by helping enterprises apply existing ones to core business tasks. Founded in 2022 by MIT PhDs Antoni Rosinol and Bernard Aceituno, the company has developed a no-code platform designed to help non-technical teams build AI agents that can interact with enterprise data and systems to automate a wide range of operational workflows.

This week, the company announced that it has raised $16 million in Series A funding. The round was led by Lobby VC and LifeX Ventures, with participation from Guillermo Rauch, CEO of Vercel, and Bob Van Luijt, CEO of Weaviate. Existing investors including Gradient, Y Combinator, and Epakon Capital also joined the round, alongside Soma Capital, True Capital Ventures, and several angels.

Stack AI began with a clear problem that large language models had become increasingly capable, but companies couldn’t easily integrate them with their internal data and workflows. “We applied with a random idea of semantic annotation for images to train autonomous vehicles mainly because of my background in robotics and Bernardo’s in optimization,” said co-founder Antoni Rosinol. “But we quickly realized we were too late; that moment had passed. What finally clicked, especially inside companies, was a low-code platform that could connect internal data sources with large language models, which were just emerging.” That pivot—from a robotics-adjacent project to enterprise AI agents—coincided with the launch of ChatGPT, which the founders credit as both a catalyst and development tool. “This company was built by AI, for AI, and powered entirely by AI,” Rosinol added.

🚀 Big news for anyone building AI agents – we’ve built the fastest way to deploy AI Agents!

In just seconds, you can deploy a pre-built Chat Assistant in @StackAI_HQ .

Think of it like spinning up your own custom 🚨ChatGPT or 🔥Perplexity, powered by any LLM you… pic.twitter.com/UuLHDNqdms

— Toni Rosinol (@RosinolToni) March 19, 2025

Since its public launch in April 2023, Stack AI has onboarded over 90,000 users who have created more than 100,000 AI agents. These agents are currently in use across Fortune 500 companies, banks, hospitals, government agencies, and universities. The platform integrates with more than 100 data sources, including SharePoint, Snowflake, Salesforce, and Confluence, and supports over 30 LLM providers such as OpenAI, DeepSeek, and Perplexity.

The company says its agents are already deployed to help generate RFP responses using SharePoint data, conduct web research to create investment memos, process insurance claims using OCR and web search, and train new staff using Confluence-based documentation. They are also used to audit sales calls for quality assurance and perform SQL-based data analytics.

Stack AI is built with enterprise deployment in mind. The product is designed to run in secure environments such as on-premises infrastructure or virtual private clouds and complies with SOC 2 Type 2, HIPAA, and GDPR standards. That foundation has enabled the company to work in highly regulated sectors like finance, healthcare, and defense. Clients such as LifeMD, MIT Sloan, Nubank, SmartAsset, and a Top 5 U.S. defense agency are among the early adopters.

“Enterprises could not integrate their data with emerging LLMs,” said co-founder Bernard Aceituno. That constraint led the company to develop a platform where business users without coding experience can build intelligent agents using a drag-and-drop canvas. These agents are composed of pre-built components that connect to company tools and databases, allowing them to perform jobs that traditionally required manual input or significant engineering support.

The new funding will be used to improve the platform’s usability, performance, and support ecosystem. Stack AI plans to make it easier for business leaders to design and deploy agents without engineering assistance. At the same time, it is investing in expanding retrieval quality, increasing model access, and enhancing deployment flexibility. The company is also building more documentation, learning tools, and community support to help organizations roll out agents across their teams.

Aceituno says that rapid iteration and feedback have been central to the product’s evolution. “The biggest difference we saw with this startup was that we could get Stack into the hands of people and get feedback very quickly,” he said. “That’s extremely helpful because until you put whatever you’re building into someone’s hands, you don’t really know.” The team, which now includes 16 members with backgrounds at Meta AI, NASA, BCG Gamma, IBM, and MIT, continues to prioritize feature development, infrastructure reliability, and user experience.

Stack AI is not trying to compete in the foundational model race. Instead, it is focused on building a practical layer that allows companies to deploy AI in ways that are specific, secure, and grounded in actual workflows. The premise is simple: give enterprises the tools to build task-specific agents that solve operational bottlenecks without requiring custom software development. 

“We’ve developed a platform that enables enterprises to create custom AI agents that can interact with various systems and be deployed for diverse tasks,” Aceituno said. “It’s not about reinventing work—it’s about making it more intelligent and more efficient.”