Working rhythm
Commits per month in 2026, across every repository and branch.
269 commits · 45 per month
Production dashboard
Lucas Tito
Products, projects, experiments and content — what comes out of the code and what comes out of the writing.
12 years of work · since 2014
29 repositories
By nature 22 code and automation repos 7 academic content repos
By authorship 20 own 9 collaborative
3 experiments run
Personal and public work only. It does not include code or intellectual property developed inside companies.
AI coverage
22 software and automation projects, by their relationship with AI.
- AI shipped6
- AI planned7
- No AI9
2 are native AI products — the product itself is AI — both built on MCP.
See AI techniques applied
Language models (LLM)
- GPT
- Claude — Opus, Sonnet, Haiku
- MCP (Model Context Protocol)
Retrieval and semantic search
- RAG
- RAG em grafo
- Query expansion
- Reranking
- Embeddings (OpenAI)
- Sentence-Transformers — BGE-M3, MiniLM
Classical AI, NLP and search
- Clustering (DBSCAN)
- PMI
- RapidFuzz
- Minimax
- Busca A*
Development stage
22 software and automation projects, by lifecycle.
- 6 in development
- 11 on standby
- 5 completed · unmaintained
Some have been alive since 2018 — like the platform from my master's dissertation, now gaining an LLM on AWS Bedrock.
Languages and tools
Languages by code volume (bytes).
See tools and platforms
Frameworks and libraries
- React
- Astro
- Flutter
- Node.js
- Spring
Platforms and automation
- AWS
- AWS Bedrock
- Cloudflare
- n8n
- Android
- Alexa
Engineering practices
Across the 22 code repositories.
- With documentation21 / 22
- With automated tests16 / 22
- With evaluation metrics6 / 22
- Coverage measured in CIexpanding
See evaluation metrics used
Model quality
- Precision
- Recall
- F1
- ROC / AUC
- RAGAS
Product and operations
- Conversão
- Sucesso / falha de envio
- Classificação de arquivos
- Aceitação do usuário (TAM)
Collaboration
9 repositories with other people.
- 9repositories
- 7collaborators
- MCP servers for AI
- Academic research
- Software
Publications and content
Works — across channels and repositories. It counts the artifact, not the repository.
- 6books published
- 4books in the press
- 12podcasts and videos
- 7scientific writing repositories
See the data as a table (screen readers and detail)
| repositories | 29 |
|---|---|
| code and automation repos | 22 |
| academic content repos | 7 |
| experiments run | 3 |
| AI shipped | 6 |
| AI planned | 7 |
| No AI | 9 |
| in development | 6 |
| on standby | 11 |
| completed · unmaintained | 5 |
| With documentation | 21 / 22 |
| With automated tests | 16 / 22 |
| With evaluation metrics | 6 / 22 |
| repositories | 9 |
| books published | 6 |
| books in the press | 4 |
| podcasts and videos | 12 |
| Python | 39% |
| Java | 19% |
| TypeScript | 16% |
| HTML | 13% |
| JavaScript | 5% |
| Go | 4% |