Warren Chan

I'm a dermatology resident with a background in AI and biomedical informatics from Stanford. I built Docora because I was spending too much time searching through clinical guidelines, research papers, and reference materials spread across hundreds of files on my computer.

Every tool I tried either required uploading files to the cloud or couldn't handle the complexity of medical terminology across different document formats. PDFs, Word docs, PowerPoint presentations, Excel spreadsheets. All with different structures, all containing information I needed quickly.

So I built what I needed. Docora uses retrieval-augmented generation (RAG) to let you search across all your local documents in plain English. Your files stay on your computer. You get answers with exact citations back to the source document and page.

Background

• •Stanford University, BS in Management Science & Engineering, MS in Biomedical Informatics. Took AI/ML coursework before the current wave.
• •Howard Hughes Medical Institute, Medical Research Fellowship. AI/ML research in skin cancer detection at Stanford.
• •Baylor College of Medicine, Medical school.
• •Dermatology Residency, where the document search problem became impossible to ignore.

Why I Built Docora

Most document search tools are built for developers. They assume you want to configure pipelines, manage embeddings, and write queries in a specific syntax. I wanted something that worked the way you'd ask a colleague: "What did the Phase 3 trial say about adverse events?"

That's what Docora does. It handles PDFs, Word docs, PowerPoints, and spreadsheets in one place, and gives you answers with exact citations. Doctors, lawyers, researchers, consultants, anyone who works with a lot of documents and needs to find specific information fast.