Sergio Ortiz · Data Science · Valencia

I build intelligent systems from real-world data.

I design and ship data products, machine-learning models, automation and reproducible systems for organisations and meaningful real-world problems.

View selected work
Visual overview of the UrbanFlow Valencia project
Visual overview of the GEMF — EXIST 2026 project
Visual overview of the UPV-EARTH project
GEMF · EXIST 20262026 · Published researchUrbanFlow Valencia2026 · Academic systemUPV-EARTH2025–26 · Research platformGenAI Maverick · Accenture Spain2025 · Challenge winner

01 · Selected work

Systems, not just models.

Each case connects the problem, the data, the technical decision and how it was made usable.

02 · Leadership · Community

Beyond the projects.

View experience

Coordinate, connect and turn ideas into initiatives.

Sigma Data Club, UPV Investment Club, peer tutoring and The Pink Force are a real part of my profile: operational leadership, communication, mentoring, social impact and organisational collaboration.

100+active students coordinated
467happlied AI programme
4leadership, mentoring and ambassador roles

03 · Capabilities · Connections

Explore the portfolio map.

Open interactive graph
Obsidian-inspiredProjects, capabilities, leadership and innovation in one network.Select a node to follow its relationships.

04 · GitHub

Recent activity.

Verified bootstrap

05 · Notes

Behind the result.

All notes
Using an LLM to translate, not to judgeImageOCRGeminiEncoderFusionp(y)
Using an LLM to translate, not to judgeMultimodal system
Research1 Jul 2026

Using an LLM to translate, not to judge

The design decision behind GEMF: turn visual and pragmatic meaning into structured evidence before supervised prediction.

Read note
A model is not finished when evaluation ends
A model is not finished when evaluation endsVerified project view
MLOps20 Jun 2026

A model is not finished when evaluation ends

What UrbanFlow adds around the model: publishing, an API, monitoring and decision logic.

Read note

06 · About

I care about what happens after the model works.

I build end-to-end data and AI systems for real organisations and real decisions. My work brings together modelling, software and automation; beyond the code, I lead university initiatives and coordinate teams.

More about me