Hexaware Collaborates on Project NANDA Fellowship to Build the Future of AI Agents

Hexaware launched applications for its Project NANDA Fellowship, an inaugural cohort of more than 10 engineers and external developers working on open-source infrastructure for decentralized AI-agent discovery, identity verification and secure coordination. The program supports Project NANDA, an MIT Media Lab-originated initiative focused on standards for an open agentic web, and complements Hexaware's agentic-AI investment in Zerovity. Priority application review closes September 19, 2026.

Analysis

This is immaterial to Hexaware’s near-term earnings and should not be treated as a monetizable AI-contract signal. The economic value of an open agent-discovery standard will likely accrue first to control points—identity, security, observability, cloud hosting and enterprise workflow orchestration—rather than to firms contributing engineering talent to open-source reference implementations. If interoperability lowers switching costs between model providers and agents, it is modestly negative over time for proprietary end-to-end platforms whose valuations assume closed ecosystem lock-in.

Over the next 1-3 months, the relevant catalyst is developer adoption rather than the fellowship itself: GitHub activity, external contributors, enterprise pilots, integration partners and any governance structure around standards/IP. Over 6-18 months, a credible cross-vendor agent identity and coordination layer could expand demand for zero-trust security and API management, benefiting PANW, CRWD, OKTA and APIM-oriented platforms, while increasing competitive pressure on point-solution agent vendors without distribution or security credibility.

Contrarian view: open standards are often marketed as decentralizing infrastructure but frequently fail to achieve enterprise deployment without a trusted managed layer. Hyperscalers MSFT, AMZN and GOOGL can capture this outcome regardless of which protocol wins because they own identity, compute, developer tooling and procurement channels. The thesis is falsified if Project NANDA fails to attract recognizable enterprise implementers or if agent frameworks converge instead around proprietary protocols from hyperscalers and major model vendors.

Key Decisions for Investors

• No standalone trade in Hexaware: treat this as a watch item until management quantifies AI-related bookings, utilization, pricing uplift or named enterprise deployments; fellowship activity alone has no valuation-bearing KPI.
• Maintain a 6-18 month structural preference for PANW and CRWD versus pure-play agent-application vendors: autonomous cross-organizational workflows raise identity, permissions, audit and runtime-security spend. Reassess if enterprise pilots show agents remain contained within single-vendor SaaS environments.
• Use GOOGL, MSFT and AMZN as the liquid expression if open-agent interoperability gains credible developer traction; initiate only following verifiable adoption milestones such as major framework integrations or enterprise consortium participation, not press-release announcements.
• Monitor OKTA for a potential second-order beneficiary, but require evidence that machine/agent identity becomes a paid enterprise category. The risk is hyperscaler-native identity absorbing the spend, which would favor MSFT and AMZN instead.