Reliant’s core product, Tabular, combines an LLM (including Llama 3.1) with proprietary models to pull structured data from scientific literature. The startup targets time-consuming literature review and multi-thousand-study data extraction tasks that typically require grad students or interns. Reliant emphasizes precision and recall, citing a zero-error result on a multi-thousand-study extraction where ChatGPT had an 11% error rate. The product surfaces extracted data in a UI that lets users edit entries and trace data back to original papers, and the team is building features to help users find where to spend their attention. The company integrates multiple machine-learning models trained on proprietary data to resolve domain-specific ambiguities. Reliant is compute-intensive and has purchased its own hardware rather than renting cloud GPUs, and it focuses on for-profit customers to ensure the technology can pay for itself.

Total raised

$11M

Funding rounds

1

Latest round

Seed

Latest activity

Aug 2024