Federica Cabrera Doglio
AI Engineer & Product · Paris
I build AI products end to end.
I find the problem, build the product, and bring the users.
drag to explore · scroll to travel
Before the universes
I turn ideas into products people actually use.
I'm an AI software and product engineer in Paris, genuinely obsessed with building. I work end to end across consumer and AI: the model, the agent, the growth engine, the interface. I've spent years in early-stage startups, founding my own and building inside them.
- I start early, at idea zero, and ship before it's comfortable.
- I learn and adapt fast, because the product teaches me what's next.
- I find what's worth building, then I build it and distribute it myself.
Everything past here is a universe I built. Scroll to travel.
Universe 01 · Now
Clarity
Founder · live on the App Store · my current focus
Get it out of your head. Make it real.
A voice-first AI planner for busy ADHD brains like mine, live on the App Store with real users. It's where I spend most of my time right now, iterating on real usage data. I've grown its audience organically through Orbit, the agent engine I built, with zero ad spend.
Universe 02 · Now
Orbit
The growth engine behind my own products
The AI-agent growth engine I built to run organic experiments for my products. This is my edge: I don't just build, I distribute. It proposes 7-day experiments and the variables to test across seven accounts, then learns from the data, all with zero paid spend.
- 7agent-run accounts
- $0paid growth
Universe 03 · Now
Rowads
Senior Product & AI Software Engineer · Paris
Hackathon winners · Cannes
At Rowads, an AI ads and content studio, I build the AI tools behind the production line. The main one is an agent that turns product shots and app demos into finished video ads, end to end, and lets the creative team iterate on them just by chatting with it.
Universe 04 · 2026
Rojo
Security for AI agents · I built the working MVP
Your agent's words are safe. Its actions are not.
Rojo runs a live AI agent against adversarial inputs and reports what it actually does: every tool call, every permission it crosses, with a verdict and evidence a developer can reproduce. I built the working MVP end to end, from the red-team engine to the one-command install into Claude Code or Cursor, and worked on discovery with the first design partners.
Universe 05 · 2025
OmnIA
Founder · AI decision engine for product teams
An AI decision engine for product teams: what to build next, why it matters, and how. I built it from real pain, sitting between Product and Tech. I've beta tested it with real users and product teams, where it's been a considerable time saver and gives data to back up decisions.
What it taught me: building was never the bottleneck, distribution was. So I built Orbit.
Universe 06 · 2024
FiNSPO
My first company · founded & engineered end to end
My first company: an AI fashion-discovery platform I co-founded and built end to end. A consumer app, a brand analytics dashboard, and a creator panel, all of it mine. It started as a college ML project and grew to real users.
- 1000+beta users
Universe 07 · Founding engineer
Order EAT
Founding engineer at a fast-growing LATAM startup
Real systems, real scale.
As founding engineer, I built the marketplace, the subscription billing, and the inventory systems that schools run on every day. It scaled across Latin America to serve millions of meals a year.
- 500+cafeterias
- 200k+families
- 10M+meals a year
- 10countries
Universe 08 · Engineering
Strike Security
Backend & data engineering · cybersecurity platform
Backend and data engineering at a fast-scaling cybersecurity startup: I built Go and Java microservices and data pipelines, and shipped the customer-facing reporting that unblocked adoption.
Everything I build is the same instinct, pointed at a different sky.
Eight worlds, one way of working: find what's worth building, build it end to end, then distribute it myself.
About
I move fast, on purpose.
I represented Uruguay at BEO World in London and placed second overall. I won my university's only 80% merit scholarship, awarded for the top entrance score among all women, and I was selected for mc² Cohort 1. This year I won a hackathon in Cannes with the Rowads team.
With no contacts in fashion, I once talked my way into Milan Fashion Week on 200 cold emails a day and a one-way bus ticket. And before my own products, I shipped production software at Order EAT and Strike Security.
- LLM agents & evals (MCP)
- TypeScript
- React / React Native
- Python & ML
- Generative video
- Growth & paid
- 0→1 product