Projetsleadbot-command-center
Outils numeriques

Leadbot Command Center

Un centre de commande FastAPI et Playwright pour l'extraction de leads Google Maps, les journaux en direct et les exports Excel/CSV telechargeables.

PythonFastAPIPlaywrightWebSocketsDockerPandasOpenPyXLHugging Face Spaces
GitHubLive DemoLive Demo: AvailableHosting: Hugging Face Spaces

Probleme

Ce qu'il fallait resoudre

Lead research is repetitive when business listings, locations, and contact details are collected manually. A small operator needs a focused tool that can start a search, show progress, and return usable spreadsheet exports.

Solution

Comment le projet y repond

I built the project around a FastAPI web service, a lightweight HTML interface, Playwright automation, WebSocket progress updates, and file export endpoints. The app is packaged with a Dockerfile so it can run on container-friendly hosting.

Fonctionnalites clefs

Ce que le projet comprend

FastAPI web app with health endpoint
Playwright-powered Google Maps scraping workflow
Live progress updates through WebSockets
Excel and CSV result exports
Docker-based deployment
Public Hugging Face Space demo

Architecture / decisions techniques

Choix d'implementation a examiner

Used FastAPI for a small backend surface with explicit health, scrape, websocket, and download endpoints.
Packaged the app with a Playwright Python Docker image so browser automation dependencies are available in the runtime.
Stored generated result files in a runtime results directory and exposed download endpoints for reviewer-friendly exports.

Validation / qualite

Comment j'ai rendu le travail evaluable

Local Docker image build completed successfully
Local /health endpoint returned HTTP 200
Hugging Face Space reports the Docker runtime as running on CPU Basic hardware

Ce que j'ai appris

Enseignement technique

I learned how to package a browser-automation workflow as a deployable web service, including health checks, runtime port binding, and a lightweight command-center UI.