Hexaware Technologies has collaborated with the Project NANDA Fellowship, a three-month, fully remote program designed to give developers hands-on experience building the open-source infrastructure behind what is being called the emerging Internet of AI Agents.
The fellowship is run by FAN, the umbrella organization overseeing Project NANDA, with Hexaware backing the initiative as part of its broader push into the open agentic web.
At the heart of the program lies one of the toughest technical puzzles facing the industry over the next decade: how autonomous AI agents can find, verify and work with one another reliably at scale. Fellows selected for the program will contribute directly to Project NANDA, an effort that began at the MIT Media Lab and is now focused on building the standards, infrastructure and reference systems needed for a decentralized network of AI agents. Their contributions will feed into areas such as agent discovery, verifiable identity and secure coordination across different organizations.
The first cohort includes more than ten fellows, made up of both Hexaware engineers and outside developers chosen from a global pool of applicants. Each participant will receive a stipend along with direct mentorship, and rather than working through simulated exercises, they will be writing code that becomes part of live, open-source agentic infrastructure. Organizers kept the group intentionally small so that mentorship stays close and every fellow gets a meaningful role in the work.
For Hexaware, the partnership serves two purposes: developing new talent in AI infrastructure engineering while deepening its ties to the open-source community driving the agentic web forward. The move builds on the company’s existing investment in agentic AI through its Zerovity platform and further strengthens its relationships with the research and developer circles working out how AI agents will actually function in real-world production environments.
Applications for the fellowship are now open to developers around the world, with priority given to those submitted by September 19, 2026.
About The Author
