Open to Product Management roles · physical AI & safety
Technical Product Manager · San Francisco

I build AI for the physical world.

From molecular biology to autonomous vehicles, I ship 0→1 products where software meets the real world, the kind where being wrong has real consequences and trust has to be earned. A product manager with the technical depth to lead these systems, who treats safety as a product problem: catch the mistakes early, close the loop, earn the trust.

View selected work ↗ My approach →
6+0→1 products shipped
20+Stakeholders aligned
50%Faster delivery cycles
30+Cross-functional team led
10K+Users reached

The operating thesis

/ 00 · Approach

Most software can be wrong and nobody gets hurt: a typo, a slow page, a bad recommendation. The products I build don't get that luxury. When AI acts in the physical world, a car at an intersection, an RNA molecule folding, a city deciding what's safe, being right is the whole job. You don't get there by promising it. You get there by building the systems that catch mistakes and close the loop, so the product earns trust before anyone has to rely on it.

P-01

Decision-grade, not noise

An AI output is only worth anything if someone will act on it. I design for that bar: systems that show their work and earn enough trust that a team will make a real decision on what they say, instead of second-guessing a dashboard nobody believes.

P-02

Lead from real depth

I have the technical depth to get inside how these systems actually work, so my product calls are grounded in what's real, not what sounds good in a deck. I can build, which means I know what's buildable, find the hard problems early, and earn the trust of the engineers and researchers I work with.

P-03

Range is the moat

The sharpest product calls come from moving between worlds: the lab and the user, the model and the market, the technical and the human. That range is what lets me see a problem whole and find the leverage other people walk past.

Selected work

/ 01 · Case files
Flagship · Current Sep 2025 – Present · Walnut Creek, CA

A fast, unified view of where AV driving goes wrong.

Advanced Mobility Group / Product Manager, AV Safety Intelligence

It started in a working session with the DMV, where I saw the agencies responsible for AVs had no real way to see or make sense of the edge cases, the moments where autonomous driving goes wrong in the real world. That gap became the product: a place for the agencies to surface problematic AV driving and review it fast.

  • Built for the people who oversee AVs: the DMV, cities, and local authorities.
  • Pulls every relevant data source into one view, turning a review that meant scattered digging into a fast, efficient one.
  • Flags problematic driving with physics-based risk scoring (TTC, RSS), then explains each incident in plain language so non-engineers can act on it.
  • Built the MVP end to end, now moving toward deployment.
MVP
Built end to end
3
Authorities: DMV, cities, local
20+
Stakeholders engaged
CASE 02 Apr 2024 – Sep 2024

Eclipsebio

Technical PM · AI-Driven RNA Drug Discovery

Diagnosed researcher churn as an ML-accuracy problem, partnered with data science to close the gap, and shipped a streamlined lab-to-analysis protocol that cut time from sample to client deliverable.

+8% model accuracy Faster sample → deliverable RNA structure prediction
CASE 03 May 2025 – Sep 2025

Independent AI Consulting

Product Manager · AI Product & MVP Development

Ran full product cycles (PRD → delivery) for 3 early-stage startups across consumer, biotech, and tech, building and shipping MVPs with Claude Code and agentic AI workflows.

50% faster cycles 3 startups Consumer · Biotech · Tech
CASE 04 Apr 2023 – Apr 2024

CSES Dev Consultancy

Founder & Product Manager · Software Studio

Founded a studio that shipped 6+ full-stack products for nonprofits and clients in one year, directing 6 simultaneous deliveries across a 30+ person team of PMs, engineers, and designers.

6+ products 30+ team Founder
CASE 05 Jun 2022 – Dec 2022

Packed (DormIt)

Product Manager · Consumer Delivery Startup

Led MVP ideation to launch across subscriptions, orders, and delivery, generating $10K+ MRR, lifting onboarding completion 15%, and cutting churn 20% with a gamified loyalty system.

$10K+ MRR 10,000+ students −20% churn

What I bring to the table

/ 02 · Capabilities
01 Product
0→1 Strategy Product Vision PRDs Roadmapping GTM Buyer Mapping A/B Testing User Research
02 Domain
Physical AI Autonomous Vehicles Safety-Critical Systems Risk Scoring (TTC / RSS) Computational Biology
03 AI / ML
Applied ML Local LLMs (Ollama) Agentic AI workflows Claude Code Cursor OpenCode
04 Engineering
Python (FastAPI) TypeScript Next.js / React SQL REST APIs Git Java C++

A bit about me

/ 03 · Background

A product manager who actually builds, drawn to the problems where AI has to be right about the real world.

I've been moving fast and getting it right for as long as I can remember: high-school valedictorian at 16 and a UCSD graduate at 20 in Bioinformatics with a Computer Science minor. I never traded speed for quality or quality for speed, and that pairing is the whole job when software has to hold up in the real world.

What's rarer than the speed is that I build what I ship, not just describe it, and I've done it across consumer, computational biology, and autonomous systems. Leadership came with that range, from running UCSD's 400+ member CSE Society to building and leading beyond campus, most recently with Clear Path. The same discipline shows up off the clock, in a 1,000-lb club total: the kind of result that only comes from not skipping steps.

Today I'm focused on the safety of physical AI, the products where a model has to earn trust before anyone relies on it. What drives me is the feedback loop, building the systems that surface mistakes early, because that's the difference between AI people trust and AI they quietly stop using. If that's what you're building, I'd like to talk.

Available · let's build

Let's build something that's right about
the real world.

khouryn77@gmail.com ↗
LinkedIn/in/nabilhannakhoury ↗
LocationSan Francisco, CA
Open toProduct Management · AI / physical-AI safety