Projets Techniques

Daily Growth Digest is an automated content curation system that generates daily strategic insights for B2B executives (CEOs, CROs, CMOs).

What It Does

• Fetches articles from 8+ RSS feeds (TechCrunch, VentureBeat, Marketing Week, SaaStr, etc.)

• Filters last 24 hours of content using Claude AI to select the 8-10 most relevant articles

• Generates strategic insights and summaries for each article

• Publishes a beautifully formatted HTML digest daily at 7:00 AM UTC via GitHub Actions

• Covers three categories: AI & Automation, Growth Marketing, and B2B SaaS

Tech Stack

• Language: Python 3.9+

• Key Dependencies:

  • anthropic - Claude AI SDK for content analysis

  • feedparser - RSS feed parsing

  • requests - HTTP requests

• AI Model: Claude Sonnet 4 (claude-sonnet-4-20250514)

• Frontend: Static HTML with Google Fonts (Lora + Inter typography)

• Automation: GitHub Actions (daily cron job)

• Cost: $0.05 per digest ($1.50/month)

Main Entry Points

1. digest_final.py - Main script with four key functions:

   • fetch_rss_articles() - Pulls articles from RSS feeds (digest_final.py:170)

   • filter_with_claude() - AI-powered filtering and analysis (digest_final.py:200)

   • generate_html() - Creates HTML output (digest_final.py:246)

   • main() - Orchestrates the workflow (digest_final.py:293)

2. GitHub Actions automation config

3. Output: output.html - Generated digest file

AI Governance Readiness Assessment is an AI-powered diagnostic tool for B2B companies (€2M-€20M revenue) to assess their architecture maturity.

What It Does

Assess their architecture maturity across 4 dimensions:

• 📊 Data & Systems

• 🎯 Strategy & Alignment

• 🤖 AI & Automation

• ⚡ Execution & Optimization

Users complete a 20-question assessment, receive real-time scoring with a radar chart, and get personalized AI-generated recommendations from Claude.

Tech Stack

• Frontend: React 18 (via CDN + Babel standalone), Tailwind CSS

• No build process - everything runs in the browser with inline JSX

• Visualization: Chart.js for radar charts

• AI Analysis: Anthropic Claude API (claude-sonnet-4-20250514)

• Backend: Netlify Functions (serverless)

• Database: Airtable for lead/response storage

• Deployment: Netlify

Main Entry Points

1. index.html (934 lines) - Single-page application containing:

   • React component: DiagnosticTool starts at line 19

   • 20 assessment questions defined at line 28

   • Claude API integration at line 204 for AI-generated analysis

   • Netlify Function call at line 511 for data submission to Airtable

   • Complete UI flow: intro → questions → email capture → AI analysis → results

2. External Dependencies:

   • Claude AI analysis

   • Data storage endpoint (Netifly)

3. Architecture: Pure client-side React app with no local build tools or package.json - all dependencies loaded via CDN.