Data scientist, currently doing an MSc in Data Science at the University of Bristol. I use machine learning and AI to solve real‑world problems, from gas‑plant maintenance to museum collections.
Machine learning, analytics and AI products. Web and app design is further down.
You write how you feel, and MoodMuse finds a painting that matches it. I captioned about 15,000 artworks from The Met’s open collection with BLIP‑2 on Vertex AI, enriched each one with emotion and colour metadata in MongoDB Atlas, and matched them against the mood detected in the user’s text. The API runs on FastAPI in Cloud Run.
I built and rolled out the maintenance dashboard for the Melchorita plant and analysed sensor data from gas plants with an international engineering team. I also automated work‑order failure reporting with Power Automate on SharePoint, and applied generative AI in the cloud to maintenance work.
Client work. I’m happy to walk through it in an interview.
Describe the day you want, like “museums and cafés in Barranco, nothing rushed”, and Pathora returns a route. GPT‑4 turns the request into constraints on time, energy and budget; the backend picks top‑rated places from Google Maps and lays them out on an interactive map you can save or share.
Classifies tracks into genres from audio features such as tempo and vocal presence. PCA compresses the feature space, logistic regression predicts the genre, and the trained model then labels the part of the dataset that had no genre. Accuracy held up with far fewer dimensions.
I’m a Data Scientist working at the intersection of data, AI and technology, currently based in Bristol.
At APTIM, I’ve worked with operational and sensor data from industrial environments, building data solutions for areas including Human Resources, Occupational Health and Logistics. I’m now also exploring how AI can support planning and decision-making, including a project focused on extracting and connecting information from SAP R/3 maintenance data.
My curiosity has taken me from industrial data to AI, from building digital products to experimenting with ideas of my own — including MoodMuse, a project that started with a painting.
I also design and build websites and apps for clients and for fun.
I’m happy to talk about data science roles and projects.