Inspiration
At the end of 2025, I lost my job. In my desperation, I started taking AI courses. When I found work as a mechanical assistant at a heavy machinery distributor, the fear of losing it pushed me to specialize further. One day I saw the workshop's reporting process: everything on paper, all information lost. That's when I understood I wasn't just there to repair machines.
What it does
TRAZA does two things: (1) Market Intelligence — monitors LinkedIn, forums and social media using Gemini to detect pain points in real time. (2) Service Reports — technicians fill a form, Gemini generates a professional HTML report automatically.
How we built it
I started with Google Forms + Apps Script using other LLMs, but migrated to Gemini by design and project logic. Today all analysis and report generation runs on Gemini. I learned TypeScript, REST APIs, and SQL on the fly — I'm an Agricultural Mechanical Engineer, not a programmer. Built entirely outside working hours, with my own equipment and API credits.
Challenges we ran into
Hardest part: getting 20-year veteran technicians to adopt the system. Technically: maintaining AI coherence across multiple services for the same client. All while doing my regular mechanic job.
Accomplishments that we're proud of
Built everything with no team and no budget. The system runs in production at a real heavy machinery distributor in Querétaro, Mexico. Three technicians use it daily. We generated 75 USD in month ly recurrin grevenue with our first external customer, atan AI costof only 0.08 per service and a 75x LTV/CAC ratio.
What we learned
Technology is easy. Changing decades-old processes is hard. A system that depends on one person is not a system. With Gemini, AI costs pennies — making industrial intelligence accessible to small workshops.
What's next for TRAZA
50 paying workshops in 12 months. Mobile app for field technicians. Collaborative failure prediction network. TRAZA aims to become the engine of industrial intelligence for Mexico.
Log in or sign up for Devpost to join the conversation.