01 / Foundations
Technology meets business
Cognizant · IIM Calcutta · Microsoft internship
Built a foundation in software, management and understanding business needs.
PGDCM, IIM Calcutta · BTech, Anna University
PRODUCT LEADER · BUILDER · 14 YEARS
I build products around people.
I turn complex customer problems into products that organizations can actually ship — across commerce, logistics, marketplaces and now AI.
Human-centered product thinking × complex systems × thoughtful leadership.
Amazon·Paytm·Snapdeal·Unicommerce·Google·Reliance
PEOPLE PROBLEM PROGRESS
From systems to products to AI.
14 years building across commerce, logistics and marketplaces.
Grounded in technology. Shaped by people.
- 2009–2012
- Jun 2012–Dec 2014
02 / Enterprise commerce
Reliance Retail
Strategy and Product Management
Connected inventory, orders and fulfillment through enterprise commerce systems.
40% increase in order-processing rates
- Dec 2014–Jun 2015
03 / Platform quality
Google
Product Expert, gTech Product Solutions
Improved ads-review tools and operational metrics supporting quality at scale.
- Jun 2015–Sep 2017
04 / Logistics platforms
Snapdeal & Unicommerce
Senior Product Manager → Associate Director
Built seller tools, a cloud logistics platform and a rider app linking operations to customers.
~INR 1.5 crore annual savings · 40% increase in delivery CSAT
- Sep 2017–Mar 2019
05 / Customer journeys
Paytm
Deputy General Manager, Product
Owned cart, checkout and orders across Paytm and Paytm Mall.
- Apr 2019–Mar 2024
06 / Launch and scale
Amazon Fresh India
Senior Product Manager → Principal Product Manager
Helped launch Fresh India, then led selection expansion across fulfillment centers.
7× selection · 16% annualized grocery sales from new selection
- Mar 2024–Mar 2026
07 / ML marketplaces
Amazon Middle Mile
Principal Product Manager, Relay Supply
Led regional transportation product strategy and launched RL pricing in Japan.
11% improvement in Japan spot-procurement costs
- Now
08 / Building with AI
CoachOS
Independent project · In development
Developing a coaching copilot focused on grounded summaries and human review.
BTech, Information Technology · Anna University, 2005–2009 / PGDCM · IIM Calcutta, 2010–2012
MY PRODUCT DNA
What I bring to the table
14 years of product building have taught me that the hardest product problems rarely belong to one function. They sit somewhere between customers, business, technology and people.
Customer Understanding
Start with customers, not features.
Go beyond what customers ask for to understand what they are actually trying to accomplish.
Product Judgment
Find the problem worth solving.
Turn ambiguous customer and business problems into clear product choices and explicit trade-offs.
Systems Thinking
See beyond the interface.
Connect customer experience with the platforms, operations, economics and technology underneath it.
Product × Technology
Bridge the two worlds.
Translate business and customer needs into technology decisions — and technical constraints back into product choices.
Influence
Build trust before you need it.
Create shared context across teams so difficult decisions become collectively owned rather than pushed through.
Execution
Make hard things happen.
Reduce ambiguity, unblock teams and drive complex cross-functional bets from idea to measurable outcome.
HOW I PRODUCT
From people to progress
The products change. The way I approach the problem doesn't.
01
Understand the People
Start with the human behind the metric. Customers. Operators. Engineers. Business teams. Understand what they're trying to accomplish, where they're struggling and why it matters.
02
Find the Real Problem
Move beyond "What feature should we build?" toward "What problem are we actually trying to solve?" Connect customer needs, business outcomes and technical realities. Make the trade-offs explicit.
03
Create Progress
Bring the right people into the problem. Create shared context. Make decisions. Unblock teams. Build trust. Then relentlessly drive toward an outcome that matters.
PEOPLE I'VE BUILT WITH
What they say about working with me
The work I'm proudest of has always been built with people, not by myself.
Guha led this initiative from PRFAQ to launch. Her planning and execution were commendable given the complex, multi-quarter technology effort spanning teams across India and worldwide.
Director, Technology
Amazon
Guha was at the forefront of some of our largest and most complex launches, bringing senior stakeholders together, navigating difficult trade-offs, and driving alignment. She combines clarity and intellectual honesty with the technical depth to quickly connect business needs with technology requirements.
Senior Marketing Leader
Amazon
Guhapriya led complex, high-impact product launches across multiple stakeholder groups, consistently elevating the right trade-offs and decisions. The boundaries between tech/product were seamless and the tech teams were highly leaning on her.
Senior Category Leader
Amazon
SELECTED WORK · CASE STUDIES
Hard problems. Human outcomes.
A few products that shaped how I think about customers, technology and product leadership.
COACHOS · CURRENTLY BUILDING
Learning AI by building it
Question
Can AI help coaches see what they might otherwise miss?
Building an AI copilot that turns unstructured coaching-session notes into grounded summaries and evidence-backed themes — while keeping humans in control of judgment.
GenAIEvalsGroundingHuman-in-the-loop
Explore CoachOSAMAZON · MIDDLE MILE
Teaching a marketplace what to pay
Question
How do you balance transportation cost against the risk of failing to procure supply?
Adapting an existing US reinforcement-learning pricing model to launch spot truck procurement pricing in Japan.
Impact: 11% improvement in Japan spot-procurement costs
AI/MLMarketplaceLogisticsPricing
Read case studyAMAZON FRESH
Breaking the grocery selection ceiling
Question
How do you dramatically increase grocery selection when your fulfillment centers can't hold any more products?
We challenged a fundamental assumption: that everything in the Fresh experience needed to arrive at the same speed.
Impact: 7× selection expansion
Customer ExperienceCommerceFulfillmentScale
Read case studySNAPDEAL
From operations to platform
Question
How do you turn fragmented last-mile operations into a scalable technology product?
Launching a cloud logistics platform and Android rider app connecting operations, riders and customers.
Impact: ~INR 1.5 crore annual savings · 40% increase in delivery CSAT
0→1PlatformLast MileMobile
Read case studyMY NEXT CHAPTER
Building with AI
After years of building deterministic products, I'm going back to being a beginner — learning AI product management the same way I've learned product throughout my career: by building.
Problem
Find a job worthy of AI
Start with the human problem—not the model. In CoachOS: help coaches create continuity from unstructured session notes.
AI behavior
Define what the model should do
Translate product intent into grounded summaries, candidate themes, structured outputs and explicit boundaries.
Trust
Design for imperfect outputs
Validate evidence deterministically and keep the coach in control of editing, approval and judgment.
Evaluation
Define and measure "good"
Build representative evals, calibrate human and automated evaluation, analyze failures and iterate.
The model suggests. The system validates. The human decides.
Explore CoachOSTHINKING OUT LOUD
Things I'm thinking about
Notes from building products, learning AI and trying to understand people better.
I was doing customer interviews wrong.
Lessons from CoachOS customer discovery: collect evidence, explore real behavior, and look for reasons your idea might be wrong.
The PM skills that matter more in the age of AI
Understanding the real problem, making trade-offs, building trustworthy products and measuring outcomes.
I’m a Traditional PM Learning AI by Building an AI Product
Part 1 of building CoachOS: what I learned about scoping, grounding and trusting LLMs
Coming soon
Building Through People
Influence, trust, engineering partnerships and getting difficult things done.
LET'S TALK
The technology keeps changing. The way I believe great products get built doesn't.
I'm interested in building products where technology solves meaningful human problems — and working with people who care deeply about doing that well.