AI systems engineer focused on building, evaluating, and integrating intelligent tools into real developer workflows.
I design and prototype AI-powered systems including RAG pipelines, MCP-based agents, model evaluation workflows, and API-driven automation. My background in documentation and developer experience gives me a strong edge in building systems that are not just functional—but usable, maintainable, and scalable.
- 🧠 Model evaluation & applied ML (HF evals, LoRA, experimentation pipelines)
- 🛠️ AI tooling (MCP, agents, structured prompting, tool use)
- ⚙️ Backend automation (API orchestration, JWT flows, data pipelines)
- 📦 Developer experience & docs-as-code systems
- Build and evaluate AI systems (RAG, agents, small models)
- Design developer-facing AI tools and workflows (MCP, API integrations)
- Implement backend automation pipelines (JWT flows, data pipelines, CLI tools)
- Apply docs-as-code principles to improve AI system usability and DX
- Ingests RSS feeds, saved URLs, and local notes
- Generates structured summaries with embeddings
- Stores memory for semantic retrieval and trend tracking
- Supports queries like: “What changed in my tracked topics this week?”
- Fully automated authentication + request flow
- Designed for integration with tools like Bruno
- Reduces manual API testing overhead
- Implementing transformer architectures from scratch in PyTorch
- Exploring tokenization, attention mechanisms, and training loops
- Designing MCP-based tools for document navigation and code-aware assistants
- Building structured prompt systems for tool-using agents
- Developing end-to-end API workflows (JWT → bearer → service integration)
- Running and analyzing model evaluations (HF evals and alternatives)
- Training and experimenting with small models & LoRAs
- Studying transformer internals (attention, RoPE, tokenization)
- Building experimental training loops and evaluation pipelines
- Automating end-to-end API workflows in Python
- Building CLI tools and internal utilities
- Creating backend services to support AI experimentation


