Hexaware Backs Project NANDA AI Fellowship

Hexaware Technologies is supporting the Project NANDA Fellowship, a three-month remote program for developers building open-source infrastructure for AI agents. Hosted by FAN, the umbrella program behind Project NANDA, the fellowship brings together more than 10 participants from Hexaware and a global applicant pool. Fellows will work on systems for agent discovery, verifiable identity, and secure coordination across organizations. Applications are open worldwide, with priority review for submissions received by September 19, 2026.

Building Open Agent Infrastructure

The fellowship focuses on how autonomous AI agents can discover, verify, and coordinate with one another at scale in agentic AI workflows. Project NANDA, which originated at the MIT Media Lab, is developing standards, reference implementations, and infrastructure for an open, decentralized Internet of AI Agents.

Participants will contribute code to Project NANDA’s open-source community rather than completing a classroom-only program. The work is intended to support live agentic infrastructure and address coordination across organizational boundaries.

Remote Program for Developers

The inaugural cohort combines Hexaware engineers with external developers selected from around the world. Fellows receive a competitive stipend, direct mentorship, and hands-on development experience during the three-month remote program.

The small cohort model is designed to give each participant close access to mentors and chances to contribute. Developers can apply through the Project NANDA Fellowship application link.

Hexaware’s Agentic AI Work

For Hexaware, the collaboration supports training for AI infrastructure engineers while contributing to the open-source community working on the open agentic web. The program also complements the company’s agentic AI investments, including its AI delivery layer, Zerovity™.

The fellowship gives developers a route into practical open-source work on the systems intended to help AI agents identify capabilities and coordinate tasks without creating bottlenecks or security vulnerabilities.