Devothama
Gundugurki
Narasimhamurthy

Ex-Mercedes-Benz R&D · Johns Hopkins MSEM · TEDxJHU Speaker · Open to US Opportunities

Framing Problems. Building Products. Devothama
Gundugurki
Narasimhamurthy

Product & Technical Program Manager · AI Products · Data Platforms · Operations

I turn complex systems into products and programs people can run: AI decision tools, data platforms, and engineering portfolios, grounded in safety-critical work at Mercedes-Benz.

Featured Work Resume ↗ GitHub ↗ LinkedIn ↗

Now building in public CareerForger · a human-approved AI job-application system I use every day →

Devothama GN — Product and Technical Program Manager

Builder · Systems Thinker · Photographer

One flagship per lane —Built.

Each lane leads with the project that best proves it. Every case study covers the problem, the decisions, and the work, and links to GitHub for the full build and docs.

01 Flagship Build · AI Systems AI PM · TPM · Product Ops · Built in Public

Flagship · In Daily Use

CareerForger

A job-application system I built and run for my own search. It turns a job description into a tailored resume and a ready-to-review application. Every resume claim must trace to real evidence, AI only writes where writing is needed, and a Chrome extension fills the form but stops before Submit. I make the final call.

PythonSQLiteChrome ExtensionEvidence-Backed AIHuman Review

Case Study → GitHub ↗

6resume families

7human-stage invites

18regression tests

Job description → screen → route → evidence → AI draft → automatic checks → package → autofill → my review
Automatic checks for page fit, format, keywords and unsupported claims before anything is packaged
Built so it cannot submit. Real form failures became named, tested guards.
Early results: 3 interview calls and 4 assessments so far, all past automated screening

02 AI Product Management AI PM · Product Manager · Trust & Safety

Flagship · AI Trust & Safety

TrustShield AI

A console that helps trust-and-safety reviewers decide which risky videos to act on first. It shows why each video was flagged and sends unclear cases to a person. I went from prototype to market research, then moved the product from detecting harmful content to helping reviewers decide what to do about it.

Scored 7 possible markets and dropped my strongest-looking one when the evidence didn't back it
Wrote a v2 product spec and a dated decision log that records each tradeoff
Working React prototype with 8 risk signals and a review queue for people

Market DiscoveryHuman ReviewRisk ScoringProduct Spec

Case Study → GitHub ↗

Flagship · AI Copilot + Evaluation

Campus Atlas

An AI assistant for university students, designed to answer only from cited sources and hand off to staff when it isn't confident. The working demo uses a scripted knowledge base, and I built a test harness to check its decisions before anyone relies on it.

The tests caught a crisis-response gap, which I fixed in the routing logic
Test cases went from 10/12 → 12/12 passing, with every handoff and citation check correct
Product strategy, user types, launch plan and roadmap, plus a working prototype

RAG DesignAI TestingGuardrailsLaunch Plan

Case Study → GitHub ↗

03 Technical Program Management TPM · Program Manager · Engineering Ops

Flagship · Portfolio Planning

LabOps Intelligence

An interactive planning tool for an engineering org that is growing faster than its lab space. It connects hiring growth to lab capacity, then to the build projects, budgets, risks and approvals needed to close the gap. Leaders see which issues need them this week.

24-Month ForecastCapacity PlanningRisk TriageGovernance

Case Study → Live Demo ↗ GitHub ↗

24month horizon

12projects at once

8labs modeled

Change growth or buffer and the forecast, lab usage and recommended action all update
Risks scored by likelihood × impact, and time-sensitive approvals flagged
Projects sorted into weekly-watch or executive-attention tiers

04 Data & Analytics Data Analyst · Data Engineering · Platform PM

Flagship · Data Platform

Cross-Continental Eco-Economic Data Pipeline

One analysis-ready dataset covering weather, food prices, energy and economic data for 5 countries. Four public sources are joined by country and month, served through a REST API, and refreshed weekly on a schedule. Built with a 4-person team at Johns Hopkins; I built the infrastructure and wrote the product spec.

PythonPostgreSQLFlask APIDockerCron

Case Study → GitHub ↗

4data sources

5countries

7API endpoints

One row per country per month, keyed on country code and month
Missing data stays visible as blanks instead of being quietly filled in
Runs with one Docker command, loads history on first start, then refreshes weekly

05 Business Systems & Operations RevOps · Sales Ops · Solutions · Product Ops

Flagship · AI + Sales Operations

Lead Routing Copilot

A system that sorts incoming sales leads. AI reads what each lead is asking for, but written rules decide where it goes. Anything unclear goes to a person with a written reason, and every decision is logged.

10 checks before any lead is sent automatically. A lead with a tricky note like "ignore all previous instructions" still stops for a person.
A threshold test found that one of my own checks never changed a result, and the console says so plainly
REST API, operations console, review queue, 47 automated tests, live demo

FastAPIHuman ReviewAudit LogRouting Rules

Case Study → Live Console ↗ GitHub ↗

Supporting · CRM Data Quality

Clipboard CRM Reconciliation

A take-home assignment. It compares a company's website against its CRM and proposes corrections, and nothing is changed without a person's approval.

35 locations checked against 121 CRM accounts
29 changes reviewed and applied
Second run: 0 new issues
Every change re-read from the CRM to confirm it saved · 49 offline tests

GitHub ↗

06 Strategy & Discovery Product Strategy · Consulting · Market Research

Healthcare Operations · Johns Hopkins Hospital

Bloomberg Children's Center

Families faced delays at intake and registration. We looked at where the process was breaking down.

Interviewed 10+ staff and stakeholders and mapped the intake and registration process
Compared practices at peer hospitals and defined 5+ operating metrics to track delays
Recommended process and family-experience changes, with a plan to put them in place

−15%delays (reported)

+22%family satisfaction (reported)

AI Music Education · Product Strategy

KaraOrchee

An AI music-learning product. The question: would people pay for it, and how should it launch?

Ran 20+ customer interviews and a 34-person survey
Compared 11 competitors to find where it could stand out
Built pricing and unit-economics models and a phased launch plan

85%willing to pay $10–20/mo

9:1modeled LTV:CAC

International Students · Discovery

NviGo

A service platform for international students. The question: which problem should it solve first?

Ran 20+ interviews with students in 5 countries
Mapped 8 critical pain points and scored each on severity, fit and viability
Picked housing as Phase 1 and set targets: CAC under $5, 60%+ retention, NPS above 50

5countries

8→1pain points → Phase 1

Outcomes that shipped —Impact.

26%

Defect Reduction

Post-release · Mercedes-Benz ADAS

84%

Validation Coverage

IDC6 Program · AURIX Platform

95%

Tool Adoption

News Corp · within 1 month

30%

Error Reduction

News Corp · NTS data pipeline

Where I've worked —Building.

Mercedes-Benz R&D India

ADAS Integration Engineer

Drove validation coverage to 84% and reduced post-release defects by 26% across the IDC6 ADAS program on AURIX platform
Improved release cadence by 2 days/cycle through structured defect triage and cross-geo alignment with German engineering teams

ADAS · AUTOSAR · CANoe · HIL Validation · ECU Integration

News Corp (NTS)

Product Development Intern

Built CSV-to-database pipeline for 75+ users, cutting data retrieval errors by 30%
Shipped 5+ prioritized enhancements based on user feedback, reaching 95% adoption within 1 month

Python · Flask · ETL · Product Requirements · Backlog Prioritization

CSV categorisation code from this internship ↗

Johns Hopkins Technology Ventures

Corporate Partnerships Intern

Mapped JHU research capabilities to industry priorities across emerging life sciences — market analysis, technology commercialization, and partnership strategy
Translated research strengths into business-facing insights for corporate partnership and commercialization opportunities

Technology Commercialization · Market Research · Business Development

Present Technologies

Product Strategist & Analyst

Mapped the U.S. soccer technology ecosystem across 6 competitive levels, evaluated 15+ tools, and defined 3 core user personas with full workflow documentation
Delivered executive-ready GTM report that directly influenced product direction and partner outreach

Product Strategy · Market Research · Customer Discovery · GTM

KaraOrchee, Inc.

Product Strategist & Data Analyst

Ran 20+ customer interviews and a 34-person survey; validated 85% willingness to pay at $10–$20/month
Compared 11 competitors and built pricing and unit-economics models (9:1 LTV:CAC) behind a phased launch plan

Customer Discovery · Pricing · Competitive Analysis · GTM

NviGo

Product Manager & Data Analyst

Ran 20+ interviews with international students across 5 countries and mapped 8 critical pain points
Scored pain points on severity, fit and viability; picked housing as Phase 1 with targets of CAC under $5, 60%+ retention, NPS above 50

Product Discovery · Prioritization · Success Metrics

Bloomberg Children’s Center, Johns Hopkins Hospital

Research & Strategy Consultant

Interviewed 10+ stakeholders, mapped intake and registration workflows, and benchmarked peer hospitals
Defined 5+ operating metrics and recommended process changes; project reported 15% fewer delays and 22% higher family satisfaction

Healthcare Operations · Process Mapping · Stakeholder Interviews

The person behind the PM —More Than Product.

Building products is how I solve problems. Photography is how I learn to see them.
Beyond product management, I document places through travel photography, explore different cultures, and share ideas through public speaking — experiences that shape how I observe users, communicate complexity, and stay genuinely curious.

📷

Photography

Capturing stories through travel and careful observation. A practice in noticing what others walk past.

🌍

Travel

Experiencing cultures that broaden perspective. The best product empathy comes from living outside your defaults.

🎤

TEDxJHU

Communicating ideas that inspire action. The discipline of a 12-minute talk transfers directly to every product pitch.

Explore More ↗ The best PMs stay endlessly curious about the world outside the product.

Framing Problems. Building Products. DevothamaGundugurkiNarasimhamurthy

One flagship per lane —Built.

Outcomes that shipped —Impact.

Where I've worked —Building.

The person behind the PM —More Than Product.

Building products and programs that solve real problems?

Framing Problems. Building Products. Devothama
Gundugurki
Narasimhamurthy