Noesis — a research copilot
A retrieval-augmented assistant that ingests academic PDFs and returns grounded answers with citations and in-line visualizations.
I'm Santino Vivas, an 18-year-old university student fascinated by artificial intelligence, scientific research, and decentralized systems — prototyping ideas that sit at the intersection of all three.
I'm Santino — 18, based between classrooms and late-night terminals. I study, read papers, sketch systems, and write code. My interest sits squarely at the intersection of artificial intelligence, natural sciences, and blockchain — three disciplines that look different but share the same underlying obsession: encoding reality so we can reason about it.
I believe the next decade of meaningful work comes from combining these fields: AI that explains scientific phenomena, decentralized systems that safeguard knowledge, and interfaces that make all of it feel human. I approach every project with a bias for rigor, taste, and a small amount of beauty.
Selective collaborations where ambition meets craft. I work slowly and carefully on a small number of things at a time.
From first idea to working demo — model fine-tuning, RAG pipelines, agents, and small models that actually solve a problem.
Smart-contract design, dApp front-ends, wallet UX, and tokenization patterns. Built with care — security and clarity first.
Research notebooks, data viz, simulation UIs — turning messy scientific workflows into clear interfaces and reproducible pipelines.
Premium landing pages and product sites — dark, editorial, highly animated. Interfaces that feel as considered as the product behind them.
Clear, opinionated essays on AI, blockchain, and where science meets code — for startups that want thought leadership with substance.
Pair sessions with other students and early builders — help shaping ideas, choosing a stack, or untangling a gnarly system.
An evolving toolbox — these are the things I use, teach myself, and rebuild my mental model around.
A rotating set of personal projects — built to learn, most still in motion. Placeholders for now; live links coming as they ship.
A retrieval-augmented assistant that ingests academic PDFs and returns grounded answers with citations and in-line visualizations.
Interactive 3D explorer for protein folds — a playground to understand structural biology through manipulable visualizations.
An encrypted, decentralized journal where entries are timestamped on-chain. Own your writing forever, unreadable to anyone else.
Exploratory protocol that verifies a model's outputs on-chain without revealing the model itself — early, speculative, exciting.
Shipping small tools at the AI × blockchain frontier while studying full-time. Writing, open source, and focused consulting on the side.
Shipped Chainlog and contributed to a handful of DAO experiments. Deepened into Solidity, Foundry, and protocol design.
Started formal CS studies — theory by day, building by night. First exposure to research-grade mathematics and papers.
First line of PyTorch. First finetune. Everything since traces back to that specific terminal session.
Open to research collaborations, side projects, and quiet conversations about AI, science, and where the blockchain fits in.