Available for selected projects

Francisco Micael

Backend & AI Engineer

I build dependable systems where backend engineering, applied AI, and thoughtful interfaces meet.

Start a conversationExplore GitHub
01About

Engineering for clarity, not spectacle.

I’m Francisco Micael, an engineer in Brazil focused on turning complex problems into useful, maintainable software. My work moves from data and backend architecture to the interface—always grounded in evidence, explicit tradeoffs, and the people who use the result.

  1. 01Useful before impressive
  2. 02Evidence over assumptions
  3. 03Simple systems that endure
02Expertise

Across the full path from idea to working product.

01

Backend systems

APIs and services shaped around clear boundaries, reliable data flows, and maintainable code.

  • Python
  • FastAPI
  • PostgreSQL
  • REST
02

AI & data

Applied AI, retrieval workflows, and data products designed for useful, explainable outcomes.

  • LLMs
  • RAG
  • Pandas
  • ML
03

Frontend

Focused interfaces that make complex systems legible, responsive, and pleasant to use.

  • Next.js
  • React
  • TypeScript
  • CSS
04

Mobile

Native Android experiences with durable local data and dependable background behavior.

  • Kotlin
  • Android
  • Room
  • Material
05

Delivery

Containerized, tested, observable software with a pragmatic path from repository to production.

  • Docker
  • CI/CD
  • Git
  • Testing
03Selected work

Systems built around real problems.

Abstract radar interface mapping dental clinic prospects
01Lead intelligence

Dental Radar

Problem
Sales teams need a better way to identify dental clinics most likely to purchase and prioritize outreach.
Approach
A data pipeline and scoring workflow turn public business signals into a reviewable, prioritized prospect list.
  • Python
  • FastAPI
  • PostgreSQL
  • Docker
View repository
Layered knowledge cards connected to an evidence graph
02Evidence-first knowledge

WikiAI

Problem
AI-assisted publishing needs transparent sources and human review to earn trust.
Approach
An encyclopedia MVP combines AI-assisted drafting with evidence, citations, and distinct contributor, reviewer, and admin workflows.
  • FastAPI
  • Next.js
  • PostgreSQL
  • TypeScript
View repository
Dark financial dashboard with gold market charts
03Market data

Financial Market Dashboard

Problem
Financial data from multiple sources is difficult to compare when signals live in disconnected views.
Approach
A production-minded dashboard retrieves, analyzes, and visualizes market data in one coherent analytical workspace.
  • Python
  • Streamlit
  • Pandas
  • Docker
View repository
Mobile trial tracker with a reminder timeline
04Android utility

Trial Reminder

Problem
Free trials are easy to forget, turning small subscriptions into surprise charges.
Approach
A private, local-first Android app tracks trial periods and schedules timely reminders without unnecessary complexity.
  • Kotlin
  • Android
  • Room
  • AlarmManager
View repository
04Contact

Have a hard problem worth solving?

I’m open to backend, AI, and product engineering opportunities. Tell me what you’re building and where it gets difficult.