creators

Vik Paruchuri: founder and CEO of Datalab

Vik Paruchuri runs Datalab, the document-intelligence lab behind the open-source Marker and Surya projects and the Chandra OCR model, and founded Dataquest.

Vik Paruchuri is the founder and chief executive of Datalab, a document-intelligence lab founded in June 2024 and based in New York. Datalab trains small, specialised models that turn PDFs, scans, forms and handwriting into structured data, and maintains the open-source projects Marker, which showed 39.8k GitHub stars today, and Surya, alongside its Chandra OCR model, released on 30 October 2025. The company announced a $3.5M seed round led by Pebblebed in June 2025 and lists Anthropic among the labs and enterprises using its models. Before Datalab he founded Dataquest, the online data-science school, in 2014 and ran it for about eight years; Dataquest's site refers to more than a million data learners and is now led by Celeste Grupman. Earlier he was a machine-learning engineer at edX and a US diplomat, and he taught himself data science partly through Kaggle competitions. He posts his open-source and product work on X and GitHub.

Specialization

document AI, OCR, open-source ML tooling

Country

United States