Software
I build tools where research meets code.
AI, bioinformatics, audio and scientific productivity. Personal projects, prototypes and the BIO-HPC group software. Here I gather what I have built, what I am building and what I want to build.
Background
From 8-bit to supercomputing
I have been programming almost all my life. I started as a kid on 8-bit machines, with a ZX Spectrum (which I still tinker with now and then). I went on with the first PCs, before the internet. I installed my first Linux distro in 1998, a Red Hat. During my PhD I dived into FORTRAN77 and scientific computing, and from there I kept jumping to each new frontier: the Cell processor, GPUs, clusters, supercomputers and even some quantum computing. That same curiosity to squeeze every new machine is still with me.
What I like to do
AI-assisted research
Turning work with papers into something more systematic: PDF synthesis, concept maps, gap detection and assisted manuscript writing. Not replacing the researcher, but expanding what they can review, connect and write.
Drug discovery and bioinformatics
Docking, virtual screening, HPC pipelines and structural analysis. I come from the fundamentals: C, C++, FORTRAN77, advanced Linux, Bash and cluster computing. I like turning lab ideas into tools others can run.
Audio and podcast
Segment recording, transcription, clips and outreach from papers. Audio as a bridge between technical research and wider audiences.
Executable second brain
Obsidian as a knowledge base processable by agents (Cursor, Claude Code). The vault not as a dead archive, but as a codebase of ideas, protocols and projects.
Projects in production or demo
Explore Labs
An AI academic platform for researchers: PDF and reference synthesis, literature concept map, manuscript review, OpenAlex integration and transcription for audio outreach. Evolving with feedback from real users.
SegmentPod
An iOS/macOS app to record long podcasts in segments, with transcription and AI suggestions. Made for long interviews without losing the thread. Personal use and on TestFlight; in development: live questions to the vault and transcript export.
BIO-HPC group software
Part of the software I develop or coordinate belongs to the BIO-HPC group (UCAM): a suite of tools for drug discovery, molecular simulation and scientific web servers.
- MetaScreener: HPC orchestrator for large-scale virtual screening
- ASGARD: automated molecular dynamics analysis (GROMACS)
- SIBILA: interpretable ML / AutoML
- TOLEDO: batch Maestro-Desmond MD
- STELLAR: fragment-based docking
- OptiPharm: 3D similarity and pharmacophore optimisation
- Web servers: Achilles (Blind Docking), BRUSELAS, DIA-DB, OBE-DB, AntiAge-DB, ConFiLiS, Vina_vision
Idea lab
What I want to build
- GAP Screener: method and app to detect scientific gaps in the literature
- Live virtual screening: interactive exploration around a hit
- MetaScreener single-command: one command, YAML config, MCP and AI-first dashboard
- Metadock / ULTRADOCK: virtual screening engine from scratch, full-GPU
- Books from notes: assembling manuscripts like code from the vault
- Podcast from paper: from article to episode with an assisted pipeline
- Knowledge OS: local RAG over my note base to think with LLMs
- Problematic-citation detector: cross your bibliography with Retraction Watch and predatory journals
My stack
Development: Cursor + Claude Code (agentic IDE) · Python · TypeScript/JavaScript · WordPress REST API · GitHub
Fundamentals: C · C++ · FORTRAN77 · advanced Linux · Bash · HPC · SLURM · Singularity · GPU/CUDA
Science: GROMACS · AutoDock Vina · PyMOL · Maestro/Desmond · RDKit · ChEMBL · OpenAlex
AI and data: OpenAI (Whisper, GPT) · Claude · MCP · autonomous agents · scikit-learn · LIME/SHAP
Environment: Obsidian (Smart Connections, Dataview) · Zotero · macOS/arm64 · iOS
From code to business
My entrepreneurial side
I have an entrepreneurial streak that runs independently from science. I enjoy the whole journey: taking an idea of my own, turning it into something that works and seeing whether it can stand on its own as a product or a business. It is a personal drive, and much of what you see on this page comes from it. That is exactly what I look for in a collaborator: building things that, besides being well made, can earn their keep.
- SaaS and tools for researchers (Explore Labs came out of this)
- Turning knowledge and experience into digital products (training, content, info-products)
- Automation and productivity with AI and agents
- Audio and podcast as a product
Shall we build something together?
If you want to start a software project with real potential and join forces (a technical collaboration, not 1:1 mentoring), get in touch.
The personal products (Explore Labs, SegmentPod) are independent from the mentoring I offer on this site and from the BIO-HPC/UCAM group services. This page gathers my technical work; mentoring has its own space.