class ParthAgrawal(AIProductManager, Educator):
def __init__(self):
self.title = "AI Product Manager · Educator · Learning Technologist"
self.based_in = "Ahmedabad, India — open to relocation (IST · PST · EST)"
self.mission = "Take frontier AI from prototype -> P&L -> people"
self.range = "8+ yrs shipping AI across EdTech, FinTech, Gov & Life Sciences"
self.school = ["IIM-Ahmedabad (AGMP '26)", "B.Tech IT, Nirma"]
self.partners = ["Anthropic", "OpenAI", "AWS", "Google Cloud"] # exec point-of-contact
self.also = "ex-Google Certified Professional Data Engineer (2022)"
def what_i_do(self):
return [
"Ship GenAI products — 30+ solutions, concept -> production",
"Turn AI capability into revenue — $250K closed, $2M pipeline",
"Teach it forward — 10,000+ engineers, 200K+ community",
"Architect enterprise AI on Vertex AI, OpenAI & Bedrock",
]AI Product Manager · Educator · Learning Technologist. I sit where the model meets the P&L — then I teach the room how to get there. A decade across EdTech, FinTech, Government & Life Sciences: 30+ GenAI solutions shipped, $250K closed on a $2M pipeline, logistics cost cut 30%, and 10,000+ engineers trained at Accenture, Pfizer, Disney & BNP Paribas.
| Revenue & Cost | Built & Shipped | Reach & Teaching | Recognition |
|---|---|---|---|
| $250K closed · $2M pipeline | 30+ GenAI solutions | 10,000+ engineers trained | 2× AWS Innovation Award |
| −30% logistics cost | 15+ production systems | 200K+ community built | 4× hackathon best-project |
| $1M saved · 500K orgs migrated | 10+ AI roadmaps | 50+ enterprise sessions | Employee of the Year ’25 |
Worked as single executive point-of-contact for hyperscaler architecture & budgeting
 |
Samsung Research — patentable work (IPS, Beacons, RCS, 5G) & real-time on-device messaging NLP |
 |
Pearson US · DeepLearning.ai · Udacity · Playpower — courses, auto-graders & adaptive learning |
 |
Early-stage startup — omnichannel analytics & adaptive recommendation engine |
 |
Mid-size agri firm — Gates Foundation-linked data work |
 |
Startups across UAE & Hong Kong — built a freight voice agent, −30% logistics cost |
 |
Elsevier · JAMA · Springer Nature — GenAI for academic publishing |
 |
NeGD (MeitY-backed) — enterprise AI across government divisions |
Languages
Frameworks · Cloud · Tooling
AI · LLMs · Agents
|
A neural net fed on my own poetry, trained to write new verse in my voice.
|
A Dialogflow tutor-bot that runs Python snippets and surfaces StackOverflow fixes in chat.
|
|
Point it at a YouTube channel, list every video, and bulk-export transcripts.
|
Optical character recognition served over a simple HTTP API — from my computer-vision days.
|
10,000+ learners across Accenture · Pfizer · Disney · BNP Paribas · Deloitte (US) · J&J · Telenor · Crown Prince Court, Dubai · 6-year Udacity mentor
- Beyond the Agentic AI Hype — Masterclass
- Knowledge Graphs & MCP for Agent Memory — Masterclass
- Prompting for Productivity — 3-hour workshop for a Fintech
- Intro to AI — guest session at IIM-Ahmedabad
- AI for Product Managers · AI for Students & Coders — Webinars
- Rebuilt Google's TensorFlow Specialization — lifted its rating 3.9 → 4.4
- "Going from 0 to 100 in AI" · "How not to land a Data Science job" — Talks
- Building GenAI revenue products & agentic workflows — MCP · RAG · Knowledge Graphs
- Sharpening the business edge at IIM-Ahmedabad (AGMP ’26)
- Exploring agent memory, local-first AI tooling & creative ML
- Ask me about taking GenAI from prototype to P&L, agent memory, and teaching AI at scale
- A neural net trained on my own poems now writes new verse in my voice —
poetry-generation-lstm
Open to senior AI Product · GenAI Leadership · AI Education roles.
