AI PRODUCT MANAGER · BANGALORE · STATUS: OPERATIONAL

Pritomsreeta Borah

Build v10.0 — Decision Intelligence & Supply Chain AI

I build AI platforms that move the needle — from a Multi-Agent Executive Intelligence Platform running at 1,000+ monthly active users to a Business Control Tower that freed $22M in working capital in a single quarter.

impact.json
$
working_capital_freed: "$0M"
monthly_active_users: "0+"
markets_delivered: "0+"
years_experience: "0+"
01 — About
Where supply chain depth
meets AI product thinking

Ten years ago I was reconciling freight manifests and hunting root causes by hand — chasing down why a forecast was wrong before it turned into a stockout. That's where I learned to read a supply chain like a story, not a spreadsheet.

That instinct followed me into AI. I translate between people who rarely agree — planners who trust their gut, engineers who trust their models, executives who just want the answer — and ship products all three can live with.

Right now, that means running two products out of one 12-person pod at UPL: a Multi-Agent Platform giving leadership a real-time view of the business, and a Control Tower that's already freed $22M in working capital. Built on LangChain, LangGraph, n8n, Lovable. Shipped, not slideware.

Get in touch ⇩ Download resume
2025 —
NOW
Product Manager
UPL LTD · BANGALORE
Active
2023 —
2025
Senior Consultant — PM
ERNST & YOUNG · BANGALORE
2021 —
2023
Product Owner
CAPGEMINI · BANGALORE
2017 —
2021
Senior Business Analyst
ACCENTURE · NOIDA
2015 —
2017
Business Analyst
TRANSPORT CORP OF INDIA · MUMBAI
02 — Stack
What I bring to every product
Agentic AI
LangChain LangGraph Multi-Agent Systems RAG Knowledge Graphs LLM Applications Speech-to-Text Text-to-Speech
Product craft
Product Strategy Roadmapping PRDs User Research A/B Testing Figma GTM Mobile App Design
Supply chain
S&OP Demand Planning Control Towers Inventory Optimization EXIM Intelligence Statistical Forecasting
Data & analytics
Databricks Power BI SQL PySpark Delta Lake Azure AI
Architecture
API Integration FastAPI PostgreSQL SSE Streaming Agent Orchestration
Tools & prototyping
n8n Lovable Figma Jira Notion Confluence
03 — Releases
5 shipped products. Real outcomes.

Currently running two flagship AI products at UPL — the Multi-Agent Executive Intelligence Platform and the Business Control Tower. Previously shipped across EY, Capgemini, and Accenture.

v3.2.0 · UPL · 2025 Shipped
Multi-Agent AI Executive Intelligence Platform
LangChain · LangGraph · Knowledge Graph · 12-person pod
Problem
Leadership was making calls across 6+ fragmented systems — EXIM, FX, commodity, market news — with no unified real-time view.
What I built
A multi-agent decision layer on LangChain/LangGraph, with voice, mobile, and a chat assistant built into the workflow.
Agentic AILangChainLangGraphKnowledge Graphn8nLovableMobile
View full case study ↗
1,000+
Monthly active users — scaled from 100 month-over-month
6+
Fragmented data systems unified
200+
Initial leadership base expanded globally
v2.4.0 · UPL · 2025 Shipped
Business Control Tower
LLM · Databricks · PySpark · 12-person pod (shared)
Problem
Planning across Sales, Marketing, Supply Chain, and Finance was siloed and manual — forecasting errors were tying up working capital.
What I built
One Control Tower unifying all four functions, with an LLM recommendation engine replacing manual forecast review entirely.
LLMControl TowerDemand ForecastingPySparkDatabricksS&OP
View full case study ↗
$22M
Working capital freed in Q4
30%
Planning exceptions cut
4
Functions unified — Sales, Marketing, SC, Finance
v1.8.0 · EY · 2023 Shipped
AI-Driven Planning Across 12+ Global Markets
Ernst & Young · 2023–2025 · Senior Consultant PM
Problem
Global FMCG clients ran demand planning on siloed manual workflows — no consistent view, chronic forecast inaccuracy.
What I built
AI-driven forecasting products and a repeatable pilot-to-scale playbook, deployed across 12+ markets with pods of ~30.
GTMUser ResearchPRDsA/B TestingFigmaS&OP
View full case study ↗
40%
Forecast accuracy improvement
30%
Feature adoption lift
12+
Global markets delivered
v1.3.0 · Capgemini · 2021 Shipped
Forecasting & Planning Product Transformation
Capgemini · 2021–2023 · Product Owner
Problem
Clients had no real-time lifecycle visibility — planning stayed reactive, teams buried in manual data gathering.
What I built
A decision-support dashboard suite with automated exception flagging, plus the first structured backlog for the planning products.
Product OwnershipDashboardsForecastingBacklog
View full case study ↗
22%
Dashboard adoption
20%
Planning efficiency up
15%
Forecast accuracy up
v1.0.0 · Accenture · 2017 Shipped
Planning & Inventory Optimization at Scale
Accenture · 2017–2021 · Senior Business Analyst
Problem
Fragmented planning across 3 hubs and 120+ locations — chronic stockouts and excess inventory happening at once.
What I built
A unified forecasting framework with variable safety stock logic and the first integrated KPI dashboards across the network.
Inventory OptimizationStatistical ForecastingS&OPAnalytics
View full case study ↗
75%
Forecast accuracy achieved
30%
Stock-outs reduced
120+
Markets optimized
Status: Open to work
Let's build something that matters

Hybrid or remote. Open to Senior PM, AI PM, and Lead PM roles.

04 — Playground
Also shipped: a tiny game.

Even PMs need a break. Arrow keys or WASD. Don't let the snake eat itself — unlike most product roadmaps.

SCORE: 0
BEST: 0