Hi, I'm Sparsh Nagpal

AI Product Manager

AI Product Manager with experience across the full product lifecycle—from ideation and user research to deployment and adoption at both fast-moving startups and Fortune 500 scale. Built and shipped LLM, RAG, and generative AI products serving 3.5M+ DAUs and $1B+ in business value.

Sparsh Nagpal

About Me

Get to know me better

About Me

I've been in AI since 2021—six years across disability tech, a mobile app development company, early-stage pre-funded startups, scaled bootstrap ventures, and giants like Boeing. I've also TA'd LLM courses at the University of Washington and built my own startup experiments along the way.

My titles have often said ML Engineer or Data Scientist, but the work has always been product-shaped: I've ideated, planned, and shipped end to end—roadmaps, user research, prototypes, and production deployments—not just models behind the scenes. That means real comfort with both 0→1 (scrappy MVPs, founder rooms, first users) and 1→100 (adoption at scale, stakeholder alignment, enterprise rigor).

The hardest part of AI today isn't learning one framework—it's keeping up with how fast the field moves. That's where I do my best work: spotting the right problem, shaping a practical solution, and knowing which tools are actually worth using this month—not last year.

What I Bring

  • Product DNA: PRD writing, A/B testing, user research, stakeholder alignment, and GTM collaboration with technical credibility
  • Gen AI Expertise: LLMs, RAG & multi-agent systems, diffusion models, prompt engineering—shipped to production, not just POCs
  • Full-Stack Thinking: Cursor/V0 prototyping, Supabase/pgvector, Figma, FastAPI, and observability dashboards
  • Startup Agility + Enterprise Rigor: Built fast at scrappy startups and scaled thoughtfully at Boeing

I'm passionate about the next wave of AI products—not incremental features, but transformative experiences that feel magical yet deliver concrete ROI. Currently exploring AI Product Management roles where I can leverage both technical depth and product instincts.

Core Expertise & Skills

Product & Prototyping

Product Roadmap
User Research
A/B Testing
PRD Writing
Figma / UI/UX
Cursor / V0

Generative AI & ML

LLMs
RAG Architectures
Agentic AI
Multi-Agent Systems
LangGraph
Diffusion Models
Prompt Engineering
RecSys

Programming & Frameworks

Python
SQL / VectorDB
FastAPI
Docker
MCP Tooling
Tableau

Professional Experience

My career journey and achievements

June 2024 - Present

Data Scientist 2 — AI Products

The Boeing Company · Seattle, WA

  • Agentic Text2SQL RAG Chatbot: Owned roadmap, UI/UX, and cohort-based feedback loops for a multi-agent copilot (LangGraph/LangChain, DeepAgents)—conversational Tableau queries, SQL-to-visualization, and document ingestion. Drove 100+ analyst adoptions across 4 business units; reduced insight turnaround 4 min → <30s; cut dashboard effort by 90%.
  • Sales Order Cancellation Prevention: Identified $1B+ in at-risk orders—owned full ML stack, engineered 60+ features, fine-tuned XGBoost, built production APIs, and layered LLM evaluation for explainability. Designed automated part substitution workflow.
  • GraphRAG for Contract Intelligence: Directed OCR model selection (Tesseract, LayoutLM, DocTron) and LLM pipeline for financial clause extraction—improving accuracy 25%, cutting manual review 30%.
  • Promoted to full-time on performance; won Boeing Hackathon 2024 (2× extraction accuracy); named Top 100 Intern in the U.S. (WayUp/Yello).
LangGraph RAG XGBoost GraphRAG FastAPI
2023 - 2025

Master of Science in Data Science

University of Washington · Seattle, WA

Graduate program in advanced statistical methods, machine learning, and data engineering with focus on generative AI and neural networks.

Advanced Statistics ML Theory Data Engineering Research Methods
June 2023 - Sep 2023

AI Engineer — Product Development (contract)

Flavor AI · Remote, India

  • Product Roadmap & Market Research: Collaborated with founders to define MVP roadmap for a GenAI ad-creation platform—competitor benchmarking, client interviews, and feature prioritization.
  • Image Generation Rebuild: Rebuilt diffusion-model pipeline and doubled generation quality using LoRA, ControlNets, and prompt engineering.
GenAI LoRA ControlNet Diffusion Models
Mar 2022 - June 2023

Machine Learning Engineer — Gen AI and Content Products

Appyhigh Technology · Remote, India

  • AI Avatar App: Built and launched Face Avatar GenAI app reaching 1k+ DAUs within one month at $12 ARPU, validating early product-market fit.
  • Diffusion AI Editor (phot.ai, 1M+ users): Managed roadmap for AI photo editing platform—background editor, object removal, avatar builder with cross-functional coordination.
  • Content Feed & Recommender SDK: Owned personalization SDK powering 15+ apps; A/B tests grew DAUs 2M → 3.5M (+30%). Built 25+ dashboards across 50M+ documents.
GenAI RecSys A/B Testing Product Roadmap
Mar 2021 - Mar 2022

ML Intern — Founder's Office

Friends for Inclusion · Remote, India

  • Product R&D: Co-designed India's first Text-to-ISL translation application; achieved 97% comprehension accuracy in user testing.
  • Backend & Deployment: Led model deployment via AWS Lambda, DynamoDB, Docker—reduced real-time response latency by 60%.
  • Research Publication: Published at Microsoft-backed IIIT-B Empower Conference, top tech-accessibility venue in India.
AWS Lambda Docker NLP Accessibility
2018 - 2022

Bachelor of Engineering in Computer Engineering

University of Mumbai · Mumbai, India

Undergraduate degree in Computer Engineering with focus on software development, algorithms, and computer systems.

Computer Engineering Software Development Algorithms Systems Design

Featured Projects

Some of my recent work

ReRout - Open-Source LLM Router

ReRout — Open-Source LLM Router

Oct 2025 – Dec 2025 | Personal Research Project · 26★

Production-grade microservices platform for intelligent LLM routing with ONNX-based intent/complexity classification via DistilBERT, PII guardrails, A/B testing, and Prometheus/Grafana observability. OpenAI-compatible drop-in API reduces inference costs 40–60%. Open-sourced with active external contributors.

LLM Routing DistilBERT ONNX PII Guardrails Prometheus FastAPI
LLM Traffic & Brand Monitoring System - Answer Engine Insights Dashboard

GeoniusAI — LLM Traffic & Brand Visibility

Oct 2025 – Jan 2026 | Personal Research Project

Built GeoniusAI, an enterprise-grade AI visibility SaaS with a proprietary Visibility Score combining AEO, GEO, and brand metrics across ChatGPT, Gemini, and other AI tools. MCP kit orchestrates 30+ SEO APIs via LangGraph multi-agent architecture with human-in-the-loop middleware. Full-stack solo execution from Figma → Cursor → Vercel.

LangGraph MCP AEO / GEO Multi-Agent AI Vercel Figma
Memorable-ai - Persistent LLM Memory Library

Memorable-ai

Dec 2025 – Present | Personal Research Project · 7★

Developing a 2-line persistent LLM memory library using Mem0 + interceptor architecture (inspired by mem0 and memori). Zero-code integration that injects relevant memories before LLM calls and extracts new memories after conversations—works with OpenAI, Anthropic, LiteLLM, and 100+ models.

Mem0 LLM Memory Python Interceptors Open Source
Meta Analysis Project

Meta Analysis & Research Platform

Built advanced meta-analysis tools using statistical modeling and machine learning to synthesize research findings and generate insights from large-scale datasets.

Statistics Research Methods Data Synthesis Python
AI Content Generation & Slogan Fine-tuning - Product imagery transformation

AI Content Generation & Slogan Fine-tuning

Created an intelligent content generation system using RAG architecture and agentic AI frameworks to produce engaging social media posts and marketing content. Also developed fine-tuned language models for generating creative and effective marketing slogans, utilizing advanced NLP techniques and prompt engineering.

RAG Agentic AI LangChain LLMs Fine-tuning NLP Content Generation
Resume Intelligence Platform

Resume Intelligence Platform

May 2025 – Oct 2025 | Volunteer — corporate training firm

Built resume intelligence tool parsing 2,000+ resumes with Gemini LLM and Supabase VectorDB, using advanced RAG techniques, hybrid semantic/rule-based matching, and Google Drive API integration at near-zero infra cost.

Volunteer project for a corporate training firm

Advanced RAG Gemini LLM Supabase pgvector Google Drive API
NetApp Data Analytics Platform - ML-Based Ransomware Detection in Virtual Machines Research Poster

NetApp Data Analytics Platform

Developed comprehensive data analytics solutions for enterprise storage systems, implementing advanced ML models for performance optimization and predictive analytics.

Capstone project, report proprietary to organization

Machine Learning Python Data Analytics Enterprise Solutions

Get In Touch

Let's collaborate and create something amazing

Email

0sparsh2@gmail.com

Send Email

LinkedIn

Connect with me professionally

Connect

Location

Seattle, WA, USA