About

Hey. My name is Gonçalo, currently living with my family in Zurich, Switzerland. I am also the Co-Founder of Prismus.ai, where I am building the the next data layer that will change how science and industry interact. And once that works well enough, build the models of the future that will transform information into knowledge. In my free time, I run, paint, and I hike as an excuse to look for rocks to crack open to see if they have any treasure inside. I also have a Bachelors and Masters in AI. I finished my bachelors with a thesis and an internship from the cardiology department at the UMCG, the Netherlands; and my Masters in Deep Learning at the Pollice Research Group of the Stratingh Institute for Chemistry, University of Groningen, the Netherlands where I started the data pipeline that turned into Prismus: consuming, transforming, and interpreting scientific data into actionable causal chains that will help us solve many of the world's most complex problems. I worked for 5 years as an AI research engineer for the Icelandic Institute for Intelligent Machines (IIIM), Reykjavík, where I became a product owner and manager of the AI team where I supervised 3 full time engineers and several interns throughout the years. Before that, I worked for 2 years for DASH, UMCG, the Netherlands, first as an AI junior scientist and then as a data scientist. While at the UMCG, I taught at the DASH Summer and Winter Schools, mentoring students and researchers in ML techniques, and headed several research projects with real medical data and medical personnel to interpret and understand the underlying mechanisms of different disease models.

I am particularly interested in recent advancements in Deep Learning and Deep Reinforcement Learning, Retrieval-Augmented Generation (RAG) techniques such as temporal and causal RAG, Multi-Agent Systems, and multimodal models like JEPA. I am particularly interested in the intersection of these technologies together with massive data and compute and how they can be used to solve real-world problems.

What excites you?

Current Roles

Co-Founder of Prismus.ai

• What is Prismus: Prismus is a scientific intelligence system that converts fragmented technical sources—papers, books, datasets, codebases, instrument manuals/spec sheets, and live industry signals—into a single evidence-grounded, time-indexed knowledge graph: each document yields a verified “micro-graph” of methods/experiments/results/claims with precise provenance pointers, these are merged into a global “macro-graph” with typed citation roles, contradictions, and supersession over time, and the graph then powers graph- and temporal-reasoning tools (via API/MCP) that let users and agents navigate science by lenses, retrieve the exact supporting evidence for any statement, and generate grounded outputs plus forecasts (e.g., which labs will adopt which modalities/instruments next and why) that update as new evidence arrives. From this moat, we can spin off products and train models, that will hopefully help us solve many problems.

Previous Roles

Data Scientist – DASH, UMCG, Groningen, Netherlands (2022–2024)

• Lead ML projects in healthcare, collaborating with medical researchers to analyze and interpret complex datasets.
• Serve on the board of the Machine Learning Community at UMCG, fostering discussions on ML applications in medicine.
• Teach at the DASH Summer and Winter Schools, mentoring researchers in ML techniques.

AI Consultant & Engineer – Itemis, Germany (2022–2023)

• Provide AI expertise for consulting projects, focusing on ML engineering and AI-driven solutions, namely Multimodal models.

Teaching Assistant

• Amsterdam UMC in Biology and Machine Learning, Netherlands (2020–2021)
• Think Tank VSNU Kennistafel, Netherlands (2019–2020)
• Assisted in courses such as Autonomous Systems, Cognitive Psychology, Multi-Agent Systems, and Computer Vision (2019–2023)
• DASH Data Science Summer School (2022–2024)
• DASH Data Science Winter School (2022–2024)

Additional Contributions

• Co-organized a symposium on Accountability, Responsibility, and Transparency in AI (~300 attendees).
• Hosted COVERTalks, inviting people such as Joscha Bach and Stephen Wolfram.
• Founded DataDump, a student initiative to teach data analytics.
• Co-created GAPC, the largest programming competition in Groningen.
• Former editor and head of content for COVER magazine, writing on AI topics.
• Board member of the Machine Learning Lab at UMCG, fostering research discussions in AI (2023–2024).
• Board member of the student association COVER for CS and AI at RUG (2020–2021).