Reliant builds Tabular, a data-extraction platform that uses an LLM (partly LLaMa 3.1) augmented with proprietary models to pull structured data from messy scientific documents. The company targets time-consuming literature-review and systematic-review tasks, aiming to replace manual extraction done by grad students and interns. Reliant emphasizes high precision and recall for domains where mistakes are costly, reporting zero errors on a multi-thousand-study extraction task described in the article. The startup is compute‑intensive and has purchased its own hardware rather than relying exclusively on cloud providers to support large pre‑extraction workloads. Founders and leaders include CEO Karl Moritz and co‑founders Marc Bellemare and Richard Schlegel. Reliant focuses on selling to for‑profit customers so revenue can cover GPU costs and support further development. The product includes a UI that lets users edit extracted data and trace entries back to source literature.

Total raised

$11M

Funding rounds

1

Latest round

Seed

Latest activity

Aug 2024