U10 · Secondary
IRS Helper — RAG over US tax documents
A RAG system over IRS forms and publications. It crawls PDFs and EPUBs from irs.gov, indexes them in pgvector and answers through an OpenAI-compatible API that OpenWebUI uses as if it were a model.
private
Engineering decisions
- Scrapy crawling: separate spiders for irs.gov PDFs and EPUBs, stored in Postgres. EPUBs are converted to PDF and uploaded to Cloudflare R2.
- Vectorization with OpenAI embeddings (1536 dimensions), skipping non-English versions of the same document.
- Two retrieval modes: when the question asks for a form ("Form 1040"), a regex pulls the form number and returns a direct link with no LLM call. Otherwise, vector search with Cohere reranking and an answer grounded in the retrieved context.
- FastAPI server compatible with `/v1/chat/completions` (including `stream: true` over SSE), so any OpenAI client works unchanged.
- Exposed in OpenWebUI through LiteLLM, which registers the server as a model named `irs-helper`.
- Web dashboard with index stats and a list of documents not yet vectorized.
Stack
PythonScrapyPostgreSQLpgvectorOpenAI EmbeddingsCohere RerankFastAPILiteLLMOpenWebUIDocker
What it shows
People who need to quickly find a form or a rule in a large set of tax documents, from a chat interface they already use.
More projects
U5
ReachantB2B lead-list builder for Brazil and the US. It finds businesses on Google Places by industry and city, then enriches each lead with website, contacts and decision-maker. Includes an API and an MCP server for AI agents.
liveU6NutriAppPractice software for dietitians: patient records, meal plans, lab results and scheduling, with each practitioner's data isolated at the database level. AI reads lab reports, with mandatory review, and drafts a consult-support report.
liveU7DockPlus AI Client WebsitesFast, search-friendly, on-brand websites for small businesses in Brazil and the US, built with a repeatable process I wrote as a Claude Code skill that goes from the client's Instagram and Google Business Profile to deploy.
live