I will build an ai parser that turns any document into markdown, text or json

D
doanvanan293
D
doanvanan293
An D.

About this gig

I parse any document into the format you need: clean Markdown or plain text for LLMs and search, or JSON with the key fields extracted to your schema. PDFs (native or scanned), images, Word files and complex Excel workbooks with tables, merged cells and multiple sheets are all fine.


How I work:

  1. You send a few sample files and tell me what output you need.
  2. I pick the approach that fits: classic OCR, a vision-language model, an LLM extractor, or a mix, and check results against your samples.
  3. You get the pipeline, an accuracy report and clear instructions to run it.

What I bring: I built document extraction that runs in production for courts and banks (95.4% field-level accuracy), and a serverless platform that processes a page in about 5 seconds for $0.002. If your data can't leave your servers, I can run everything on self-hosted models.

Get to know An D.

An D.

Full Stack AI Engineer

  • FromVietnam
  • Member sinceOct 2026
  • Languages

    English, Vietnamese
I'm a full-stack AI engineer who ships AI features to production: document parsing and data extraction, RAG chatbots and AI agents over your own data, and voice assistants, plus the backend, frontend and AWS setup around them. I've built document extraction for courts and banks (95.4% field accuracy), a policy Q&A agent used by 1,000+ employees, and a serverless parser at about 5 s and $0.002 per page. Two RAG papers accepted at SoICT 2026 and ICTIS 2026. Send me a few sample files and I'll tell you what accuracy and cost to expect.

My Portfolio