I will build a python rag document assistant with source citations

A
andrii_tynskyi
A
andrii_tynskyi
Andrii Tynskyi

About this gig

I build small Python document assistants that search Markdown files and provide answers with source citations.

Depending on the package, I can set up SQLite FTS5/BM25 retrieval, connect an OpenAI-compatible language model, and create a command-line demo with setup instructions.

The retrieval in this service is keyword-based (BM25), not vector or embedding search. Answers depend on the supplied documents and language model, so they should be checked against the cited sources. My portfolio includes DocuAgent, a document assistant with retrieval, citations, and offline evaluation.

Please message me before ordering. Share a few non-confidential sample Markdown documents, example questions, and your preferred model/API. Never send passwords or API keys through Fiverr messages.

Get to know Andrii Tynskyi

Andrii Tynskyi

Junior AI Software Engineer

  • FromPoland
  • Member sinceOct 2026
  • Languages

    English, Ukrainian, Polish
I’m a Junior AI & Software Engineer and Computer Science student at ZUT Szczecin. I build practical backend, web, and AI solutions using Python, Java, PHP, Spring Boot, SQL, and REST APIs. My projects include a document-search assistant using RAG, a machine-learning prediction pipeline, an LLM evaluation tool, and a cinema booking API. I can help with API development, database-backed applications, Python scripts, data processing, and AI prototypes. I focus on clear communication, reliable delivery, and code that is easy to maintain.

My Portfolio