Applied AI & Knowledge Graph Engineer @ Institute for Systems Biology
Polyglot engineer building AI/ML systems, data pipelines, and backend services across a range of languages. My focus area is biomedical knowledge graphs: I build the Tablassert ecosystem, which turns messy tabular data into NCATS Translator-compliant knowledge graphs, including an autonomous agent pipeline that derives table configurations with GEPA prompt optimization. I also contribute to the open-source Translator stack (biolink-model, Babel, RTX, shepherd).
- Languages: Python, Rust, Go, Elixir, SQL, Shell
- AI/ML: Deep learning (PyTorch), LLM agents & MCP (LangGraph, LangChain), model fine-tuning (Transformers, PEFT), gradient boosting (LightGBM), classical ML & NLP (Scikit-learn, spaCy, sentence transformers)
- Knowledge Graphs & Ontologies: Ontology mapping & modeling (Biolink, LinkML, KGX), graph construction & analysis (Neo4j, NetworkX), analytical & relational stores (DuckDB, Polars, PostgreSQL, Redis), ETL pipeline design
- Backend & Infrastructure: API services (FastAPI, Gin), Nix/NixOS & nix-darwin, Docker, virtualization (QEMU/KVM), CI/CD (GitHub Actions)
- Biomedical Domain: Multi-omics, clinical-trial data, biomedical entity resolution & knowledge integration (NCATS Translator / Biolink)