Built from work this engineer published themselves. It makes no claim that they are available for hire, and it is marked noindex until they claim it. Source: sh-reya.com. This is you, or it should not be here? Write to hello@aiengineer.me.
ai-engineer
Shreya Shankar
DocETL, and research on operationalising LLM pipelines over messy data.
PhD researcher at UC Berkeley working on data management for LLM pipelines, and lead author of DocETL, a system for declarative document-processing pipelines with LLM operators. Publishes on evaluation practice, prompt iteration, and the distance between a working demo and a pipeline that holds up on real documents.
Technical stack & depth
Evaluation
Evals, traces, and a quality bar that is not a vibe check.
Retrieval / RAG
Indexing, chunking, reranking, and grounding answers in a corpus.
Production hardening
Latency, cost, observability, and the path from a notebook to a service.
Engineered solutions
- Repositoryucbepic/docetlgithub.com
Lead author of DocETL, a declarative system for LLM-powered document-processing pipelines.
- Write-upWriting on LLM pipelines and evaluationsh-reya.com
Public research notes on evaluation, prompt iteration, and data quality in LLM systems.
Something here out of date or wrong? Write to hello@aiengineer.me and it will be corrected or removed.