LIVE QUERY CONSOLE

Ask about my work

Grounded on this site’s content — a small demo of the AI/LLM work described above it.

Answers are AI-generated from this site’s content — verify anything important with Simon directly.

DRAWING SL‑2026Systems & full-stack engineering

Simon Laborde

Backend & full-stack engineer with an AI / LLM focus, built on a foundation in enterprise backend systems and full-stack development.

I bridge the gap between complex systems and the people who depend on them — currently building LLM-backed tools, taking freelance and contract work across Europe, and applying to master’s programs.

Available for contract & freelanceApplying for Fall 2026 master’s — MoSIG, Sorbonne, PSL, TUM, RWTH

Start a conversationDownload CV (PDF)
Drawn by
S. Laborde
Origin
07°13′S 35°53′W
Bearing
Brazil → Europe
Revision
2026.5
Sheet
1 / 1
Scale
1:1
  • PT Portuguese — native
  • EN English — C2
  • FR French — C1
  • ES Spanish — B1

Journey

One throughline: learning how systems behave, then making them serve real needs — from physics engines to research infrastructure.

  1. 2024 — present

    Freelance Software Development

    Independent — remote

    Built LABNOV's bilingual site for a UFCG materials-research lab — a Sanity CMS content model for its non-technical staff and an automated publication pipeline off Brazil's Plataforma Lattes — and NutriQuest, a gamified nutrition-tracking app. Additional client work under NDA.

  2. 2020 — 2025

    B.Sc. Information Systems

    UNIFACISA — Campina Grande

    Thesis — “Applications of Generative AI in Clinical Decision Support” — defended with distinction, and the seed of the MedHelp project.

  3. 2022 — 2023

    Software Developer

    Accenture

    Optimized billing and backend processes on Oracle BRM with C and shell scripting; automated financial reporting through Oracle BI Publisher inside an Agile delivery team.

  4. 2018 — 2020

    Game development foundation

    UNIFACISA

    Interactive systems design, C# gameplay logic, and physics-engine work — where the instinct for how systems move and break first took hold.

Technical arsenal

Read as a legend, not a tag cloud — grouped by the layer of the system each part touches.

Languages & runtimes

Python · Java · C · C# · TypeScript · JavaScript · PL/SQL

AI / LLM

LLM integration & prompt systems · Model Context Protocol (MCP) · agent & subagent design · Claude / Anthropic APIs · NLP · scikit-learn · pandas · NumPy

Backend & data

RESTful APIs · PostgreSQL · MySQL · Oracle BRM · Oracle BI Publisher

Frontend & mobile

Next.js · React · React Native · Vue · Expo · Tailwind · semantic HTML/CSS

Infrastructure & tooling

Docker · AWS EC2 · Vercel · Git · JIRA · Sanity CMS · Figma

Testing & load

pytest · JMeter

Credentials

Anthropic — AI Fluency: Framework & Foundations · Anthropic — Building with the Claude API · Anthropic — Introduction to Model Context Protocol · Anthropic — Model Context Protocol: Advanced Topics · Anthropic — Claude Code in Action

Selected work

Four projects, four problems. Each drawing is styled after what the thing actually is. scroll →

  • 01 · AI clinical decision support

    MedHelp

    01Exam-recommendation output02NLP normalisation layer03PEC clinical-note input

    An NLP tool for Brazil’s public health record system (PEC): it normalizes clinical notes, adapts language per audience, and flags preventive exams from patient demographics and history.

    Terminology accuracy
    92%
    Recall
    100%
    Model surface
    NLP

  • 02 · gamified nutrition tracker

    NutriQuest

    01Mobile UI · quests, streaks, XP02Offline-first local store03Background sync engine

    A cross-platform mobile app that turns nutrition tracking into a game — daily quests, streaks, and a competitive leaderboard, fully localized for Brazilian users.

    Level 7 · Nutritionist-in-training1,840 / 2,400 XP
    1. 1ana.map3,120
    2. 2you1,840
    3. 3tiago.rn1,610

  • 03 · bilingual research platform

    LABNOV

    01Bilingual Next.js site02Sanity CMS content model03Automated Lattes sync

    Site bilíngue para o laboratório de pesquisa LABNOV, com sincronização automática de publicações a partir da Plataforma Lattes e edição de conteúdo via Sanity CMS para a equipe não-técnica.

    O toggle acima é o mesmo componente do site real.

  • 04 · enterprise backend, Accenture

    Enterprise Billing Optimization

    01BI Publisher reporting02Batch rating & billing (C)03Oracle BRM platform

    Backend performance and reporting-automation work on an enterprise billing platform — profiling and tuning batch processes and replacing manual reporting steps with scripted pipelines.

Detail 04.1 — LABNOV

The Lattes sync pipeline

A close-up on one part of the LABNOV build: keeping a research lab's publication list current, automatically, past an obstacle that rules out full automation.

  1. 01

    Plataforma Lattes

    Brazil's academic-CV registry. No public API; the export is CAPTCHA-gated.

  2. 02

    reCAPTCHA checkpoint

    Can't be automated. The pipeline pauses for one human click, then runs unattended.

    manual step

  3. 03

    Parser

    Pulls publications from the export; normalises authors, venues and years.

  4. 04

    Diff

    Compares against the CMS — new papers in, nothing duplicated or dropped.

  5. 05

    Sanity CMS

    Only the changes are written, into the schema staff already edit by hand.

  6. 06

    LABNOV site

    New publications show within a sync cycle. Nobody maintains the list.

Problem
A public publication list that drifts out of date the moment someone forgets to update it.
Approach
Automate everything except the one step that can't be — and make that step a single click, not a chore.
Result
A near-zero-maintenance list that stays honest to the source of record.

Languages

Multilingual and Europe-bound — measured on the CEFR scale, read like a signal meter.

Contact

Open to master’s program conversations, freelance, and contract work in Europe.