I build production-grade systems, contribute to high-impact open source projects, and document what I learn about AI-native engineering, agents, infrastructure, and developer tooling.
I’m focused on the intersection of enterprise software engineering, AI-native systems, and independent research.
Most AI demos work in notebooks. My interest is different: building systems that can survive real-world constraints — security, scale, observability, deployment pipelines, legacy integrations, and production failures.
Right now, I’m actively exploring and contributing around:
- 🤖 Agentic workflows, RAG pipelines, and autonomous software systems
- 🧠 AI-assisted developer tooling and infrastructure automation
- 🐍 Python ecosystem reliability, packaging, testing, and documentation
- ☁️ Cloud-native platforms across AWS, GCP, Kubernetes, Docker, and Terraform
- 🔐 DevSecOps, CI/CD security gates, compliance automation, and observability
- 📚 Research-driven engineering: turning experiments, failures, and patterns into useful frameworks
- ✍️ Weekly LinkedIn posts on AI engineering, open source, and hands-on POCs
This section is automatically refreshed from GitHub and shows recently merged PRs authored by me.
| Project | Merged Pull Request | Merged |
|---|---|---|
| pytest | Fix strict options from addopts | 2026-05-08 |
| NumPy | BUG: exclude pycache directories from wheels | 2026-05-07 |
| PyTorch | [Docathon] Convert tensor_view.rst to MyST Markdown | 2026-05-07 |
| Excalidraw | fix(editor): prevent duplicate lasso toolbar item | 2026-05-06 |
| pandas | DOC: clarify missing-value handling in pandas and NumPy reductions | 2026-05-06 |
I prefer contributions that are small, testable, review-friendly, and useful to real maintainers.
I’m also building research credibility around autonomous systems and AI-native engineering.
- Instruction Strategy Design for Autonomous Machine Learning Experimentation Systems
Read on Sciety
Small fixes compound.
Clear tests build trust.
Good documentation scales knowledge.
Production discipline makes AI useful.I like working on issues where the solution is not just code, but a clean loop:
- Reproduce the bug
- Understand the maintainer’s intent
- Keep the fix minimal
- Add targeted tests
- Explain the impact clearly
- Share the learning publicly
- Practical bug fixes in respected open source projects
- AI engineering experiments and agentic workflow POCs
- Backend and platform automation examples
- DevSecOps, CI/CD, testing, and infrastructure notes
- Weekly learning logs connected to my LinkedIn posts
- Research notes on autonomous systems and AI-native software design
I use LinkedIn as a public engineering journal: what I fixed, what I learned, what maintainers care about, and how AI changes the way we build software.
Recent themes:
- Picking better first issues in high-signal repositories
- Writing PR descriptions that maintainers actually want to review
- Debugging Python, ML, and developer tooling issues
- Turning small merged PRs into credible public proof of work
- Building AI-era engineering habits without losing production discipline
I’m always interested in conversations around:
- Open source contribution strategy
- AI-native engineering and agentic systems
- Platform engineering, DevSecOps, and cloud automation
- Production-grade RAG and internal AI assistants
- Building a public technical brand through real shipped work
