DBTB INT Richard Downe r
summary
Casetext uses cheap, scalable compute to run older language algorithms blisteringly fast and ship an extractive summarization pipeline that pulls concise, readable passages from judicial opinions. The claim is concrete: summaries cut reading time and highlight the holdings and reasoning so users see what to read and what can be skipped. Demos show the system running at scale and improving legal research workflows rather than remaining a prototype.
AI-written summary · about
related talks
40mJudgement Day: Benchmarking "Black Box" LLMs With Open Legal Datasets
22mEmpowering journalism-approaches to automatic text summarization
35mTowards Interpretable Data Science
31mFrom Data to Precision: Fine-Tuning LLMs for Proprietary Query Languages
24mHarnessing the Power of LLMs in Advanced Analytics
28mFrom Tables to Answers: building QA System for In-Document Searches
37mTaming the Data Tsunami
07mdbverse: Composable database libraries for larger-than-memory scientific analytics