
Voidscape — Teaching AI Agents to Watch Video
An AI coding agent can read images and PDFs — not video. Voidscape decomposes any video into frames plus a transcript, and prices the whole job before spending a cent or a token.

An AI coding agent can read images and PDFs — not video. Voidscape decomposes any video into frames plus a transcript, and prices the whole job before spending a cent or a token.

A business with no website is a web developer's best prospect. FindLeads searches Google Places for them, flags the website-less as tier-1 leads, and wraps a tiny CRM around the results.

Toronto has a beloved open-source tech map. Peru didn't. Peru Grid maps 53 researched startups, consultancies, incubators and funds across Lima and Arequipa — in one dependency-free HTML file.


AI-powered recruitment platform: public career pages, resume parsing, drag-and-drop pipelines, role-based admin, and the email/storage/webhook infrastructure a real hiring team needs.
Full-stack engineer, currently focused on the frontend, building scalable interfaces with React, Next.js, and Node.js. I've shipped AI-driven apps and engineered systems end to end.
Now going deeper into AI infrastructure and high-performance engineering. Three disciplines I move between, usually on the same project: engineering, AI, and data.
Full-stack with a frontend focus. I turn designs into responsive, event-driven interfaces and back them with modular auth, reusable components, and CI/CD pipelines.
I wire LLMs into real products, like AI that parses resumes into structured, filterable data, and automate the busywork around them with agents, RAG, and webhooks.
Data-driven by default. I build ETL pipelines that process millions of records and turn them into Power BI dashboards and reporting that replace hours of manual work.