I built a tool that finds company data + summarizes websites automatically (with LLM fallback)

Hey everyone 👋

I’ve been working on a small automation tool for scraping and analyzing company data, and it turned out way more useful than I expected.

Here’s what it does:

pulls company data from DNB (name, industry, website, etc.)
visits the company website and extracts clean text
finds phone numbers across multiple pages (/contact, /about, etc.)
generates a short business summary (3–5 sentences) using LLMs
handles rate limits + model fallbacks automatically

The goal was simple:
👉 turn a raw list of companies into structured, enriched data with minimal manual work.

Stack:

Python (requests, BeautifulSoup, pandas)
Groq API (multiple models with fallback logic)
retry + anti-rate-limit logic for APIs

What surprised me:

phone extraction is harder than it looks (formats are chaos)
LLM fallback logic is critical if you process at scale
scraping “clean text” from websites is still messy in 2026 😅

Use cases I’m exploring:

lead enrichment
market research
quick company profiling at scale

I also put together a simple landing page:
👉 https://dibara512.github.io/my-site/

(MVP version — focusing on functionality first)

Curious if anyone here built something similar?
Or has ideas how to monetize this kind of pipeline?

Happy to share more details 👇