AAYUSH

@ Deloitte

Built a fairness validation and audit platform for education-loan risk models, enabling transparent and accountable AI decision-making.

• •Designed a GenAI-powered validation pipeline to evaluate bias across 3,000+ synthetic user profiles covering diverse demographic scenarios.
• •Identified hidden algorithmic biases across five key dimensions: geographic, income, gender, credit, and edge cases
• •Developed policy-driven mitigation checks with before/after score comparisons to ensure auditability and compliance.
• •Introduced a Human-in-the-Loop feedback system, allowing auditors to iteratively refine model behavior and improve fairness metrics.
• •Delivered a dashboard for non-technical stakeholders, translating complex fairness signals into concise, actionable insights.

@ Airtel

Worked on production-grade backend systems and internal platforms in a high-scale telecom environment.

• •Engineered a Voicebot Onboarding platform, reducing customer integration time by 27%.
• •Fine-tuned Speech-to-Text models for Indian languages, improving transcription accuracy in real-world scenarios.
• •Built scalable architectures (Next.js, Spring Boot, MongoDB) supporting 10,000+ concurrent requests with sub-200ms latency.
• •Integrated REST APIs with LLM fallback handling, reducing manual intervention by 40% and improving intent recognition by 25%.
• •Strengthened system reliability through robust validation and request/response optimization.

Led technical initiatives and operations for a university-wide AI/tech community.

• •Mentored student teams on project design, ML implementation, and technical storytelling.
• •Spearheaded hackathons and workshops, framing problem statements and managing end-to-end execution for 500+ participants.
• •Oversaw budgeting, sponsorships, and financial planning to sustain large-scale community programs.
• •Conducted core team recruitment, evaluating both technical and leadership potential.