Skip to content

I run an AI organisation inside a Fortune 5. I still build.

I lead enterprise AI at UnitedHealth Group. 57+ engineers and researchers. Multi-million dollar P&L. 80+ products in production.

Senior AI/ML Engineer by title. Organisation, architecture, delivery and governance in scope. Research engineer by trade.

Yash Sharma, standing, in a dark suit and white shirt

Scale.

Organisation
57+ engineers and researchers
32 direct · 25 indirect · grown from 4
Economics
Multi-million dollar budget, full P&L
Eight-figure vendor contracts negotiated
Enterprise AI
80+ products in production
Strategy · architecture · production · governance
Outcomes
$37.2M+ combined annual savings
3,200 bots · 43M+ authorizations a year · 10M+ daily API calls
Governance
AI/ML Review Board, founded and led
Standards · Responsible AI · model risk
Venture
Founding Technical Partner, Optum & UHG Ventures
Startup evaluation · investment advisory

How I lead.

  1. Build when ownership compounds.

    Licensing, lock-in, security and cost decide build versus buy.

    Snowboard retired an eight-figure licensed estate: 3,200 bots at 99.7% uptime, no per-bot licensing.

  2. Bound the risk. Then ship.

    Put the security boundary where the regulator needs it and let the teams move behind it.

    CAGE: 15 teams run AI-written code inside it, 78% fewer threats, 40% lower cost of ownership.

  3. Start with the economics.

    Cost per decision, workflow and adoption before the choice of model.

    MNM: a 6.75M-parameter model deciding 43M+ authorizations a year instead of a frontier model call.

  4. Central standards, local execution.

    One review board sets standards and release gates. Forward deployed pods do the shipping.

    The AI/ML Review Board I founded, and CRYO as the shared framework every core line of business builds on.

  5. Measure in production.

    Uptime, throughput, adoption and savings are the scorecard. Demos are not.

    117,808 decisions a day, 10M+ daily API transactions across 10PB, $37.2M+ combined annual savings.

How I got here.

4 → 6 → 12 → 21 → 57+ engineers. Five roles at UnitedHealth Group since January 2019, a year at DRDO's Centre for AI and Robotics before that.

  1. Sep 2025 - Present

    UnitedHealth Group (Optum)

    Senior AI/ML Engineer

    Team
    32 direct reports (12 engineers, 20 AI/ML) + 25 indirect
    Budget
    Multi-million dollar budget with P&L responsibility
    • Founded and lead AI/ML Review Board for enterprise governance
  2. Feb 2024 - Sep 2025

    UnitedHealth Group (Optum)

    Lead AI/ML Engineer

    Team
    21 direct reports (12 engineers, 9 AI/ML) + 18 indirect
    Budget
    Multi-million dollar budget
    • Chaired AI/ML Review Board for enterprise governance
  3. Mar 2022 - Feb 2024

    UnitedHealth Group (Optum)

    Data Scientist - R&D

    Team
    12 direct reports
    Budget
    Multi-million dollar budget
    • Developed synthetic data models and digital twin simulations for HIPAA compliance
  4. Jul 2019 - Mar 2022

    UnitedHealth Group (Optum)

    Associate Data Scientist

    Team
    6 direct reports
    • Created synthetic datasets mirroring real-world properties for 20+ operational teams
  5. Jan 2019 - Jul 2019

    UnitedHealth Group (Optum)

    Software Engineering Analyst

    Team
    4 engineers (Lead Engineer)
    • Applied ML models optimizing inventory and distribution in healthcare facilities
  6. Jan 2018 - Dec 2018

    DRDO (Centre for AI and Robotics)

    Research Trainee Intern

    Clearance
    Classified Security Clearance
    • Smart Realtime GIS System: Automated detection of objects' speed, orientation, and properties

Governance, boards and venture.

AI/ML Review Board, UnitedHealth Group

Founding Member & Lead R&D Judge

Founded and led. Architectural standards, Responsible AI guidelines and model risk management.

Optum & UHG Ventures

Founding Technical Partner

Startup evaluation, investment advisory, portfolio mentorship.

Enterprise compliance

Mar 2022 - Feb 2024

SOC 2 Type II, ISO 27001 and GDPR programmes. Model risk management frameworks.

Technical foundation.

U.S. patents
213
75% deployed in production
Research publications
700+
First author on majority
AI World Championships
Consecutive wins

Patents licensed by five of the biggest names in big tech and deep tech. Topics: Large Language Models, Transformers, Healthcare AI, Computer Vision, Responsible AI, Quantum ML.

Recognition

  • 6× Consecutive AI World Champion

    Dell, Apple & NVIDIA AI Championships · Multiple Years

  • Top Patent Holder

    UnitedHealth Group · 2024

  • Industry Speaker & Panelist

    Snowflake, Adobe, NVIDIA Conferences · 2024-2025

Let’s talk.

Executive roles, board and advisory work, venture diligence, speaking. I reply within 48 hours.