Professional work, an independent public prototype, and supporting design notes. Each entry identifies its context and evidence boundaries.
Carta Healthcare · Data Scientist
Built clinical ETL and NLP pipelines at Carta Healthcare and applied LLM-assisted extraction to registry workflows. A retrospective on data preparation, workflow boundaries, and evaluation considerations.
TripleTen · Data Science and AI/ML Instructor
Data-science project review, technical feedback, live sessions, and Python/ML instruction at TripleTen. Educational work with an emphasis on clear reasoning and evaluation assumptions.
Independent Build · AI Systems Builder
Built an independent RAG evaluation prototype with 12 synthetic-adversarial documents, 22 labeled queries, and three lexical retrieval baselines, separating retrieval scoring from deterministic answer evaluation.
Design note
A conceptual design note on defensible validation, model comparison, and error analysis. No public implementation, deployment record, or reproduced evaluation result is linked here.
Design note
A conceptual note on PDF retrieval, chunk boundaries, grounding, and failure analysis. It is separate from the source-backed public RAG prototype and does not establish measured reliability.