The 20 AI agents that run my workday
How I built a team of small AI agents that handle most of the planning, forecasting and reporting behind a large AI hardware program, and what changed once they were running.
Read the post →Medhavi Chandra
Product & Program Manager · San Francisco Bay Area
I build products people use — and the systems that ship them.
I started my career answering customers' questions at a young investing startup in Bangalore. Listening closely to them led me into product, where I built apps used by more than a million people, then to an MBA in Boston, and now to Silicon Valley, where I run part of AMD's AI hardware program, with AI agents I built doing much of the coordination.
9+years in product & programs
1Musers in year one (Niyo)
$100B+AI infra program scale (AMD)
20+AI agents running every day
01 My story
My career has taken me from a customer-support inbox in fintech startups to the AI infrastructure behind some of the world's largest compute deals. The roles have changed a lot. The way I work hasn't: listen to customers, let data decide, and build the tools that help teams move faster.
-
2017 · Bangalore
Starting where the customers are
My career started on the front line. After a commerce degree at Delhi University, I joined an investing startup to run customer experience, and hearing from customers every day taught me what they really wanted. I built a loop that turned their feedback into shipped features and cut response times from six hours to one. Within three years of starting my career, I was leading a vertical and managing a team of three, the youngest manager my boss knew of.
-
2020 · Fintech
Crossing into product
Spending so much time with customers made me want to build for them, not just support them. I moved into product without a technical background, as the only woman on the team, and relied on the same habit: listen first. My research steered a new digital bank toward investing instead of expense tracking, and the wealth app we launched reached a million users in its first year. That work earned me the Star Performer award.
-
2021 · Lending
Designing for how people feel
Next, at a student-loan company, I learned that good products are as much about feelings as features. Weekly conversations with applicants showed they were anxious, not just busy. Rewriting the application around those worries raised sign-up conversion 4×.
-
2022 · Boston
Going back to school to level up
By then I had shipped products, led a team and co-authored two Harvard Business Review case studies, and I wanted to get better at leading people, teams and products, and at using technology to stay ahead of the curve. That took me to Babson for an MBA in Business Analytics & Machine Learning, where I finished in the top 1% of my class.
-
2025 · Silicon Valley
From apps to AI silicon
After graduating, I took time to find the right fit and volunteered for a cause I care about along the way. I found that fit at AMD, running a chiplet program in the MI450 "Helios" AI platform, a world away from consumer apps. My instincts still applied: I saw how much time teams spent on planning and reporting, built AI agents to do it for them, and helped keep the program ahead of schedule.
How I work
👂
Customer-obsessed
I start with what people actually need, whether that's a million users of a consumer app or the engineering teams I support at AMD today.
📊
Data-driven
Even small decisions get tested. Once, an A/B test showed me a feature was hard to find not because of where it sat, but because of how it was described. The data often points somewhere you didn't expect.
🚀
Entrepreneurial
I take ownership of whatever I work on, as if it were my own company. Back in Bangalore, I even ran a dog-boarding business on the side for four years.
🛠️
AI-driven builder
I have more than 20 AI agents running across my work and personal life, from program planning and reporting at AMD to everyday tasks at home.
02 Experience
-
Apr 2025 — Present
Program Manager · AMD
Santa Clara, CA
- I own the budget, staffing, schedule and delivery for a chiplet in AMD's MI450 "Helios" platform, the rack-scale AI system behind compute deals with Meta, OpenAI and Anthropic worth more than $100B.
- I built around 20 AI agents with MCP integrations and custom skills that automate most of the program's planning, forecasting and reporting. They pulled the schedule in, reduced the need for contractors and freed me to take on more responsibility.
-
Dec 2023 — May 2024
Product Manager Intern · PTC
Boston, MA
- I shaped a $20M projected-ARR opportunity by bringing engineering, marketing and leadership together on one go-to-market strategy for a Spatial AI product.
-
May 2023 — Dec 2023
Product Manager Intern · Clasp
Boston, MA
- I shipped an ROI calculator on schedule and hit every Q4 milestone by running sprints around clear KPIs.
- The calculator attracted more than 50 B2B prospects after launch.
-
Aug 2021 — Jul 2022
Product Manager · Leap Finance
Bangalore, India
- I raised sign-up conversion 4.4×, from 9% to 40%, by redesigning banking-partner onboarding around API-based auto-fill.
- I started a weekly Voice of Customer hour and used it to rewrite each step of the application around applicants' real worries. Help text in the credit section alone lifted its completion rate by 5%.
- I launched a refinance roadmap that generated $32M in loans in its first quarter, 6% above target.
-
Jul 2017 — Aug 2021
Product Manager · Head of Customer Experience · Niyo (Goalwise)
Bangalore, India
- I launched a wealth-management app in three months, and it reached 1 million users in its first year.
- I built an engagement and retention program with marketing, projected to deliver $5M in ARR, and was named Star Performer.
- I earned my license as a Certified Financial Advisor and built comprehensive financial plans for customers, managing more than $200M in assets under management.
- As Head of Customer Experience, I cut first-response time from 6 hours to 1 while keeping satisfaction above 90%, and set up a loop that told customers when their requests shipped.
03 Cool things I've worked on
My AI job-search agent
Three Claude agents and one shared Sheet that find fresh Product Manager roles, tailor a resume and cover letter for each, and fill in applications, stopping before Submit so I make the final call.
Read the case study → View the slide deck ↗Answering money questions as a Certified Financial Advisor
I've written answers on Quora about financial planning, investing and mutual funds, drawing on my years as a licensed advisor.
104K+ views566 upvotes65 shares
Read on Quora →
Naukri.com (A): The Business of Internet in India
A case I co-authored on Info Edge, the first Indian internet company to list on the stock market. It follows Naukri.com's 30-year journey, its strategic pivots and expansion, and is used to teach entrepreneurship at business schools.
View case →Naukri.com (B): Corporate Venture Capital in India
The companion case on Info Edge's investments, from its own classifieds like 99acres and Shiksha to its corporate venture stakes in Zomato and Policybazaar, and how corporate and financial venture capital differ.
View case →Predicting loan defaults
For my MBA I built a machine-learning model that predicts whether a loan will default, with 90% accuracy. The project placed me in the top 1% of my class.
04 Education
Babson College — F.W. Olin Graduate School of Business
MBA in Business Analytics & Machine Learning (STEM-designated) and Entrepreneurship · 2024
Finished in the top 1% of my class, recognized for a machine-learning model that predicts loan defaults with 90% accuracy.
Delhi University
Bachelor of Commerce with Honors, First Division · 2017
05 Beyond work
I'm married, and my husband and I are proud parents to Raahi, a Cavapoo who runs our home. Dogs have always been a big part of my life: back in Bangalore, I missed my own dog so much that I started a dog-sitting business and looked after nearly 80 dogs over four years. These days, weekends are for family and friends, and we travel whenever we can.
🐶 Dog mom
Raahi, our Cavapoo, is the boss of the house and the best part of every day.
🎲 Board game Fridays
Every Friday is game night, with favorites like Catan, Wingspan, Azul and Critters Kitchen. I'm competitive, so I play to win.
✈️ Travel
We travel whenever we can. Iceland is my favorite place so far.