Disclaimer I work at Oracle. This article is not official Oracle communication and does not represent the company’s views. It’s my personal attempt to understand Oracle as a case study. Recently, I watched a video about Oracle’s history and realized I didn’t actually have a calm, structured mental model of the company. I knew the […]
There’s a very practical question behind many AI conversations with customers: “Do we actually have the infrastructure to run and iterate on computer vision models reliably?” This project started as a hands-on way to answer that question with a concrete artifact:a working YOLO-based image classification model, trained on a real dataset, packaged with a Gradio
When you’re running Large Language Models (LLMs) or other heavy AI workloads, one of the first big questions is:“What GPU shape do I actually need in OCI Data Science?” This guide walks you through the key concepts, the GPU shapes available in OCI Data Science, and what shape matches the most popular AI models today.
Choosing the right GPU shape in OCI Data Science for your AI Model: A practical guide Read More »
In the past weeks I’ve been working on an Oncology Forecasting prototype built with Oracle APEX + Autonomous Database (ADB). The goal is straightforward (but very powerful): let business users define a scenario (country, indication, time horizon, epidemiology assumptions, products…), run the forecast, and then explore the results. The real challenge is not only computing
A rigorous recap from a curious amateur — and what the evidence suggests we should watch in 2026 (not financial advice). I’m not a finance professional. I’m simply someone who’s curious about how the world works — and who learns best by turning headlines into a structured, source-backed narrative. After watching a Spanish video that
2025 Was a Reminder That Markets Don’t Move in Straight Lines Read More »
A common question when using Oracle Machine Learning (OML) on Autonomous Database / Autonomous AI Database is: “What exactly do LOW, MEDIUM, and HIGH mean. And when should I use each one?” These are not informal labels. They are predefined database services (service names in your wallet / tnsnames.ora) that map to resource consumer groups.
Managing ML Performance in the Database: A friendly guide to LOW / MEDIUM / HIGH on OML Read More »
Pull Requests are where good engineering habits either level up… or quietly get replaced by “LGTM, ship it.” You open a PR. You do the responsible things: Then your reviewer asks: What happens if the uploaded filename contains weird Unicode? And you realize the truth we all share: PR reviews are human-scaled. Attackers are not.
I let two AIs review my PRs: One breaks my fastAPI code, the other fixes It Read More »
I am writing this in English and publishing it on Medium and LinkedIn because what is happening in Spain with housing should not stay a local, private frustration. It is a systemic problem in a European country that likes to present itself as modern, attractive and “open to talent”, while quietly making decent housing almost
When Housing Becomes a Luxury Tax: Why Young Spaniards Are Giving Up on Their Future Read More »
How a bird-sighting anomaly project turned into a reusable H3 geospatial demo. When you work with geospatial data, you quickly run into this question: I have latitude and longitude… now what? This article is a practical walk-through of a Jupyter notebook that answers exactly that, using H3 and Madrid as a playground. The notebook originally
From Lat/Lon to Hexagons and Neighbourhoods: Learning H3 with Madrid Read More »
Lately my feed is full of posts claiming that TOON cuts LLM token usage by 30–60% and that JSON is basically “legacy bloat”. As someone who spends her days helping customers bring AI into production at Oracle, I was curious… but also a bit skeptical. So instead of arguing in the comments, I opened a
JSON vs TOON: Is your data format really the problem? Read More »
