Launch HN: Parsewise (YC P25) – Reason Across Documents with an API
Not a pilot. Parsewise lets you extract schema-compliant data from thousands of PDFs—with word-level citations across documents.
Why it matters
YC-backed startup launches API for enterprise document reasoning, solving the 'how do we validate AI extraction results' problem that Claude and GPT struggle with at scale. Uses exhaustive search instead of RAG, achieving SOTA on grounded reasoning benchmarks.
The key facts
9 to knowParsewise (YC P25) launches document reasoning API
Beats Claude Fable on Databricks OfficeQA benchmark for grounded reasoning
Exhaustive value search across documents vs. RAG sampling approach
Word-level citation lineage for all extracted values
Model-agnostic; best results with Gemini for visual reasoning
Founders: Greg (ex-Palantir, classical ETL + AI workflows), Max (ex-Bain, financial data analysis)
Use cases: insurance PDFs, transcribed calls, emails, multi-document reasoning
Deployable in private networks
Focuses on 'human harness' (verifiability/validation) over model optimization
Go to the source
Hacker Newsnews.ycombinator.com
Publisher excerpt: Hi all, it’s Greg and Max, founders of Parsewise here Parsewise transforms a bucket of unstructured data into schema compliant data retaining lineage for values resolved across documents. Imagine giving Claude a bunch of files and asking for a CSV or JSON output. If you have tried this, you know…