From MIT to outperforming tech giants, Reducto is redefining how AI processes unstructured data, and it’s a testament to how Indian-origin founders are solving critical bottlenecks in the global AI infrastructure! This is exactly the caliber of innovation we seek to highlight at GIA.
What is the news?
- Reducto, founded by MIT graduates Adit Abraham and Raunak Chowdhuri, has raised a $24.5M Series A led by Benchmark, bringing total funding to $32.9M.
- The startup has developed state-of-the-art vision models that read documents “like humans do,” transforming complex documents into AI-ready inputs with unprecedented accuracy.
- Their technology has proven to be significantly more accurate than offerings from AWS, Google, and Microsoft, by margins exceeding 20% (in some cases) in benchmark studies.
- The company has already processed hundreds of millions of pages for clients like Airtable and Scale, with one customer (Legora) reporting 3x processing speed and 90% reduction in engineering time.
Why is it interesting?
- Reducto tackles a fundamental challenge in AI adoption that causes hallucinations when processing unstructured data like PDFs and spreadsheets.
- With nearly 80% of enterprise data locked in unstructured formats, Reducto’s approach unlocks massive value for companies implementing AI solutions.
- The founders’ journey exemplifies how deep technical expertise developed at companies like Google and Nvidia positioned them to identify and solve a critical market gap.
- Their success in outperforming tech giants demonstrates how focused startups with specialized expertise can create substantial value in the AI ecosystem.
Read more: Reducto Raises $24.5M Series A Round to Help Enterprises Unlock Unstructured Data