I'm a Technical Builder focused on prototyping, experimentation, and product-market fit. I trained as an engineer and stayed hands-on, because while being in the weeds is difficult, I find the signal tends to be in the weeds - that is, the weeds of the data, the product, or the user experience. I love to build, not just for building sake, but because it inspires a better strategy.

• safe-youtube — An open-source YouTube wrapper that gives parents the parental control tools they actually want. Built it because I needed it. My kids don't use an iPad without it.
• Habit.am — Habits dont matter, only one habit does: introspection. Habit.am is a mental health reflection tool that started pre-AI and evolved into my first Applied AI product that helps you reflect through simple multiple choice questions that lead to insights, reflections, mantras and more. 
• leash — A (vibecoded) mobile bridge that connects your cloud harness to your local Pi agent.
• tardigrade — An open-source AI agent that drives a real browser to audit any web app like a user and score it against rubrics, with a screenshot behind every finding. Works with any OpenAI-compatible model.
• ant-simulation — An experiment in computational biology: a browser simulation where up to 300 army ants are each an independent LLM agent, blind and navigating by pheromone gradients. Built to push high tokens-per-second (TPS) inference, since every ant decides its move on every tick.
• pi-container-links — A Pi extension that maps container file paths to host paths and renders them as clickable OSC 8 hyperlinks in a Sources widget.
• language-tutor — A personal PWA for translating text and automatically generating flashcards to build vocabulary over time.
• co⸺span — TBD - Coming soon

• Engineering: Full-stack & Cloud (AWS, GCP). Backend infrastructure + web frontends (Next.js, React). I build a lot of PWAs
• Product: Scoping AI products, multi-modal user research, continous discovery, shipping lean MVPs, and building intuitive on-ramps for non-technical users.
• Data Analysis: Python, charting, and data communication. It's not enough to understand the data you have, you also have to know what the data you don't have could be saying.
• Agentic Coding: Setting up dockerized agents, claws, experience with many models (K2.7, GLM-5, Opus-line, OpenAI Models, Gemini offerings, Image generation, etc)

• The process changes the product. I wrote about this here.
• Not all outcomes are measurable, especially negative externalities. I wrote about this here.

• I'm most active on X at @NoemiTitarenco
• Sometimes I blog over at noemititarenco.com/
• Looking to collaborate on: Open-source, AI tools, EdTech, Mental Health Tools.
• Ask me about: Implementing AI in your teams, product strategy for niche markets, and behavior-driven design.
• Contact: noemititarenco.com/contact

⚡ Fun Fact: I once drove from LA to Fairbanks, Alaska, and back!