Vishnu Nanduri, PhD

I help organizations turn AI ambition into AI that runs in production.

I started as a reinforcement learning researcher and professor, then moved to industry to build AI products and the teams behind them. Today I run the ASEAN AI Innovation Lab in Singapore.

NowHead of AI and Innovation, ASEAN and Korea, Kyndryl

Vishnu Nanduri on stage at GovInsider AIxGov beside a screen reading AI Ambition to AI Operations
GovInsider AIxGov, Singapore, 2026

On stage at

  • Google Cloud Next
  • Microsoft AI Tour
  • GovInsider AIxGov
  • Data Centre World Asia
  • Spark Summit
  • INFORMS
  • IEEE Power and Energy Society

In the press

  • HRM Asia
  • iTnews Asia
  • GovInsider
  • Twimbit
  • Asian Business Review
  • Kyndryl newsroom
  • Corner Magazine PH
  • CRN Asia
  • Digital News Asia
  • Singapore Economic Development Board

What I believe

Things I have said in public and still stand behind.

  • The model is maybe 10% of the problem.

    An agent that works in a demo still needs data, integration, deterministic boundaries, and people to run it before it works in production.

    GovInsider AIxGov, 2026
  • AI speeds up modernization. It will not fix broken foundations.

    Generated code that compiles can still be wrong. And if nobody understands the old system well, the project stalls anyway.

    iTnews Asia, 2026
  • Agents that act on their own need enforceable limits.

    That means policy as code, clear decision rights, and human oversight. Without them, agents stay stuck in pilots.

    Frontier Enterprise, 2026
  • AI is embedding faster than workforces are adapting.

    Companies are rolling out AI faster than they are changing how work gets done. Roles have to be redesigned around it.

    HRM Asia, 2026

Career

The short version. The longer one is on the About page.

  1. Lab and regional leadership

    I led the launch of Kyndryl's ASEAN AI Innovation Lab in Singapore in June 2025 with Google Cloud and Digital Industry Singapore. Today I help Kyndryl's clients across ASEAN and Korea put agentic AI into production.

  2. AI strategy

    At Prudential I put many predictive models into production. At Munich Re I owned AI and machine learning strategy across Asia. Then I led insurance AI at Accenture and helped build the global AI practice at Kyndryl. On stage I keep making the same argument: start with the outcome, put governance in early, and build for production.

  3. Product building

    I built distributed reinforcement learning on Spark at Impetus and IT operations analytics at IBM. At Munich Re I led an image-based motor claims product, and at Kyndryl I lead teams building agentic systems for clients.

  4. Team building

    I have built and led data science and engineering teams at Impetus, IBM, Munich Re, and Kyndryl, across Southeast Asia, Japan, Korea, and India.

  5. Teaching

    I was a tenure-track professor at the University of Wisconsin-Milwaukee, where I won the College of Engineering teaching excellence award. Later I taught online at the University of South Florida as a professor of practice. Now I teach AI in financial services and responsible AI at NUS.

  6. Research

    A PhD in reinforcement learning and game theory, applied to electricity and carbon markets. My 2007 IEEE Transactions paper on market power is still cited in research today.

Around the world

I have spoken in Singapore, Seoul, Paris, Buenos Aires, San Francisco, and plenty of places in between. I have lived and worked in India, the United States, and Singapore, led teams across Southeast Asia, Japan, Korea, and India, and worked on projects in Tokyo, Beijing, and Mexico.

World map of cities where I have spoken, across Asia, North America, Europe, and South America. Projects and work in Tokyo, Beijing, and Tecate, Mexico. Lived and worked in India, the United States, and Singapore. Seoul, South KoreaAmes, United StatesBengaluru, IndiaCharlotte, United StatesJakarta, IndonesiaKuala Lumpur, MalaysiaMelbourne, FL, United StatesPittsburgh, United StatesSan Francisco, United StatesWashington, D.C., United StatesAustin, United StatesBandung, IndonesiaBangkok, ThailandBoston, United StatesBuenos Aires, ArgentinaColumbus, United StatesDenver, United StatesDetroit, United StatesKharagpur, IndiaLas Vegas, United StatesMiami, United StatesMinneapolis, United StatesNew Delhi, IndiaParis, FranceSan Diego, United StatesSan Juan, Puerto RicoSeattle, United StatesShanghai, ChinaTampa, United StatesSingapore, SingaporeTokyo: projects and workBeijing: projects and workTecate: projects and work
Spoken inProjects and workLived and workedHome base today: Singapore

Where I have spoken

Selected sessions

More than 100 talks, panels, and lectures, from INFORMS and IEEE to Google Cloud Next and the Microsoft AI Tour.

  • Sep 29 to 30 2026

    Data Centre World Asia (Singapore Tech Week)

    Are Your Facilities Ready for AI Workloads?

    With Amit Agrawal (Techno Digital), Travis Deutmeyer (ABB). Moderated by Virat Patel (Pioneer Consulting Asia-Pacific)

    PanelSingapore
  • May 5 2026

    GovInsider AIxGov 2026

    From AI Ambition to AI Operations: why agentic AI that works in a demo dies quietly in production

    TalkSingapore
  • Mar 2026

    Microsoft AI Tour Seoul

    From AI assistants to agentic, AI-native architectures

    TalkSeoul
  • Jun 24 2025

    Launch of the Kyndryl ASEAN AI Innovation Lab

    Opening the lab with Digital Industry Singapore and Google Cloud

    With Philbert Gomez (DISG), Andrew Lim (Kyndryl), Mark Micallef (Google Cloud)

    LaunchSingapore
  • Apr 2024

    Google Cloud Next '24

    Taking GenAI from prototype to production, responsibly

    With Manoj Palaniswamy

    TalkLas Vegas
  • Jun 2019

    Munich Re 5th Regional CEO Summit

    Turning Insurance Data to Revenue Using AI and ML

    TalkSingapore

All talks

The ASEAN AI Innovation Lab

Kyndryl opened its ASEAN AI Innovation Lab in Singapore in June 2025, with Google Cloud and Digital Industry Singapore. The lab helps enterprises and government agencies get agentic AI out of the pilot stage and into production. I lead it.

The lab's work is built on the Kyndryl Agentic Framework. Most of the effort goes into data, integration, the hard limits on what an agent is allowed to do, and the people who will run it.