Projects
Not case studies. The story behind each project, told from the commit history: what I was trying to do, what broke, what I changed, and what I'd do differently.
After the June 2026 earthquakes in northern Venezuela, help requests and missing-person reports were scattered across improvised sites. HelpVenz puts asking for help, offering it and finding people in one place, with no sign-up. Built with a friend from Munich and live the day after the quakes.
Read · 4 min
Agencies spend the first hour of every project asking clients things their website already answers. AskOnce is an AI onboarding agent that reads the site first, asks only what's missing by chat, voice or phone, and writes the brief. Built solo at the TUM.ai Makeathon, 3rd place in the HappyRobot challenge.
Read · 4 min
Project ideas tend to show up away from the keyboard and are half gone by the time you sit down. SpecTalk lets you describe a project out loud through Meta Ray-Ban glasses or AirPods: a voice agent called Gervis asks up to three questions, writes the spec and hands it to a coding agent. Built solo in six days.
Read · 4 min
Making one piece of content means juggling six disconnected tools. Idealizer puts images, video, voice and copy on one infinite canvas, with an AI agent that can see and build on all of it. Built solo at the Cursor AI Hackathon Hamburg; it won two tracks.
Read · 3 min
Playlists recommend songs that already exist; nothing writes one about your actual day. TuneTools turns your news, weather and calendar into a new ~60-second song each day, with lyrics from an LLM and vocals from the open YuE model on a rented GPU. A two-person build for the Kiroween hackathon.
Read · 3 min
Stats sites give League of Legends players numbers, not advice. RuneSight, built solo for AWS Rift Rewind, puts five Bedrock agents next to your match history to tell you what to fix and how you compare.
Read · 4 min
Language teachers spend hours writing practice exercises, marking them and tracking which topics each class struggles with. LangDrill generates the exercises with an LLM, grades the answers and turns the grades into per-class statistics. Built with Sebastian over a summer, then picked up again a year later.
Read · 4 min
Who wrote this
I'm Javier Peres, an Electrical and Software Engineer from Venezuela, living in Munich. I build with AI at work and outside it.
More about me