India, Mar 5: In a rigorous national competition run under India’s AI Mission, HyperVerge was selected as one of only two winners from over 25 competing teams. The win earns them INR 1 Crore in a two-year contract to bring its face recognition technology to national public examinations.
The IndiaAI Face Authentication Challenge was launched by the Government of India to identify secure, scalable, and highly accurate Artificial Intelligence solutions capable of preventing impersonation and duplicate identities in large-scale public applications. With millions of candidates applying for government positions annually, ensuring a fair, merit-based process free from fraud is a national priority. The winners were declared at the India AI Impact Summit, held in New Delhi.
Solving a National Scale Problem
HyperVerge’s winning submission addressed the core problem of application duplication and candidate impersonation. The company demonstrated its capability to deliver:
-
Automated Duplicate Detection: Identifying individuals applying multiple times using different photos, regardless of variations in lighting, pose, image quality, or age across millions of applications over more than a decade.
-
Population-Scale Scalability: Performing one-to-many face authentication against millions of identities, within authorized, purpose-limited workflows, with sub-second latency.
HyperVerge’s robust engineering and its ability to identify duplicates from low-quality application images were the critical capabilities in the modern digital landscape that helped them win the challenge. All these deployments are designed to comply with applicable data protection laws and purpose limitation requirements.
Proven Technology and Commitment to Fairness
HyperVerge’s selection follows a rigorous evaluation process where the technology was benchmarked against top industry competitors. This recognition builds on an established track record. In 2024, its face recognition technology was independently benchmarked against U.S. Department of Homeland Security standards, which is one of the most demanding evaluations in the world.
“National-scale identity systems require more than high accuracy. They require consistency, bias control, and resilience against adversarial attacks. Being selected as a winner of the Government of India’s National AI Mission challenge is a strong validation of the depth, scale, and rigor of our face recognition and de-duplication technology,” said Kedar Kulkarni, CEO at HyperVerge. “We’re honored that our technology will now support critical public sector use cases in high-integrity institutions that conduct national public examinations, where reliability directly impacts millions of citizens.”
Deep Experience in Public and Private Sectors
HyperVerge brings extensive experience to this mandate. The company currently powers identity verification and fraud prevention for major financial institutions, including leading banks and financial institutions.
HyperVerge is also a known thought leader in Ethical AI. The company’s research team recently published findings on “The Impact of Racial Distribution in Training Data on Face Recognition Bias” at WACV 2023, underscoring their commitment to building AI systems that are fair and robust across diverse demographics.