Humata AI summarizes and answers PDF questions

Stanford biology graduate and two-time entrepreneur Cyrus Khajvandi struggled to balance scientific research with his daily workload. Khajvandi developed an AI platform to summarize and answer questions about documents, particularly scientific studies, after realizing he wasn’t alone and AI technology was becoming more accessible.

Former Labelbox founder Dan Rasmuson became CTO of Humata AI in February. It quickly grew to millions of users, processed tens of millions of pages of files, and raised $3.5 million from Google’s Gradient Ventures, ARK invest, and M13.

In an email interview, Khajvandi told , “Our mission at Humata is to empower people and organizations to make smarter and faster decisions by asking questions across all their files.” The author compares Humata to ChatGPT for all files.

Humata is very easy to use. As advertised, the platform lets users ask questions about their PDF files and get answers. Khajvandi says customers include academics, law professionals, oil and gas industry professionals, and customer support. Users can upload one or more PDFs and ask questions across them.

Current chatbots like ChatGPT and Anthropic’s Claude analyze files. Khajvandi claims Humata is stronger due to its limited functionality and focus.

“People can ask AI any question and get the answer from their own data instantly with highlighted references,” he said. The recent advances in AI allow every worker to get instant answers to their questions.

AI isn’t always the best summarizer. Fast Company tested ChatGPT’s ability to summarize articles and found that the model often got content wrong, left pieces out, or invented facts.

The privacy issue is also present. Companies and users may be wary of uploading sensitive documents to Humata’s platform for processing.

Khajvandi says Humata trained its summarization models on “diverse datasets” and “rigorously tested” them for bias. He claims that Humata collects only “necessary data” and has “strong safeguards” to prevent unauthorized access.

“We ensure informed consent, helping users understand what they’re agreeing to,” Khajvandi said. We avoid inferring sensitive information without permission as our AI systems improve. For enterprise readiness, Humata follows legal and ethical standards across regions and cultures.

Khajvandi claims Humata has thousands of paid customers and plans to use the $3.58 million it has raised so far ($3.58 million, including a pre-seed round) to improve its AI, user experience, and market reach.

Khajvandi said they raised now because there is a growing demand for efficient, AI-driven solutions to synthesize insights from massive enterprise files. The funds will help us develop new features, refine our products, and expand into new markets by empowering businesses to make better and faster private data decisions with Humata.