Lorenzo Leão Dotto

Full-Stack Developer (previously Intern and Innovation & Contracts Assistant) · Sirtec

July 2025 – May 2026 · Brazil

Electricity distribution contractor with 4,000+ employees. I started in the innovation team and finished as the developer dedicated to Helix, the field-team app.

• Started the Helix project and worked across the whole app: a web app (PWA) for managing field crews, used by around 500 people working in crews of 5.
• Implemented real-time tracking with Supabase Realtime: when a crew starts an activity, managers see it immediately, and every field activity is now measured in real time.
• Offline-first architecture: with no signal, the app stores each task's start and end times locally and syncs when the connection returns.
• Integrated the crews' third-party safety-camera software into the app, bringing safety and productivity tracking into one place.
• Earlier, in the innovation team: automated 20+ spreadsheets (reports that took ~3 hours now run in under 10 minutes), built 10+ Power BI dashboards and migrated TOTVS ERP tables with millions of rows to Parquet using Python.

Stack: React, Next.js, TypeScript, Node.js, Supabase (PostgreSQL, Realtime), PWA. In the innovation team: Power BI, SQL, Python, Excel.

Lembra

Offline spaced-repetition flashcard app

React 19 · TypeScript · Vite · ts-fsrs · Dexie · Capacitor

Built to memorise hard English vocabulary. Android app in closed testing on Google Play.

• FSRS scheduling (ts-fsrs), a daily limit in minutes and new cards paused automatically when the backlog grows.
• Local-first: data in IndexedDB (Dexie), installable PWA and an Android app via Capacitor.
• Pure domain layer (scheduler, study queue) separated from the UI, with 110 unit tests (Vitest) and end-to-end tests (Playwright); GitHub Actions builds the Android app.

Plan Whisperer

Hackathon: Winner (Challenge 8), Gold Coast Open Data Challenge 2026

TypeScript · Node.js · LLM (Qwen) · Ollama · vLLM · Tailscale

An AI assistant that uses the city's open data and planning rules to help Gold Coast residents understand what they can build, with sources cited in every answer. Team of 6.

• Contributed to the backend (LightRAG service and chat API) and the deployment.
• Researched and analysed the challenge's open datasets to choose which ones the project would use.
• Fixed bugs during development.