BigQuery Release Intelligence 🚀

An automated ingestion, synthesis, and intelligence engine for Google Cloud BigQuery release notes. Built for Google Cloud Engineers (CEs), Solutions Architects, DBAs, and Data Engineering leads.

---

Highlights

• 📡 Live Atom XML Ingestion: Directly ingests the official Google Cloud BigQuery Atom feed (https://cloud.google.com/feeds/bigquery-release-notes.xml) with local caching and TTL management.
• 🏷️ Granular Item Breakdown: Parses release drops into structured items tagged by category (Feature, Change, Fixed, Security, Announcement, Deprecated) and launch stage (GA vs Preview).
• 🧠 Dual Summarization Engine:
  • Google Antigravity AI (agy -p): Synthesizes releases into Executive Briefs, Technical Architecture & Migration Guides, Changelog Digests, or ad-hoc custom architecture queries.
  • Analytical Rule Engine: Instant offline fallback clustering updates by thematic area (AI/ML, SQL Modeling, Drivers, FinOps).
• 🖥️ Interactive Web Dashboard: Premium dark-mode interface with glassmorphic accents, real-time KPI metrics, search, filtering, and one-click Markdown copy/export.
• ⚡ Full-Featured CLI: Query, summarize, inspect stats, or launch the web server directly from terminal.
• 📦 Zero External Dependencies: Built entirely with Python 3 standard libraries and vanilla HTML5/CSS3/ES6.

---

Quick Start

1. Requirements

• Python 3.10+
• (Optional for AI synthesis) Google Antigravity CLI (agy) installed and in PATH. If agy is not present, the app seamlessly uses its built-in analytical engine.

2. Run the Web Dashboard

python3 main.py --serve --port 8080

Open your browser at http://localhost:8080.

---

CLI Usage

View Summary Statistics

python3 main.py --stats

Fetch Recent Updates

# Fetch updates from the last 14 days
python3 main.py --fetch --days 14

# Filter by category or stage
python3 main.py --fetch --days 30 --stage GA --category Feature

# Search by keyword
python3 main.py --fetch --search "JDBC"

Generate AI Summaries

# Executive briefing (Last 30 days)
python3 main.py --summarize --days 30 --mode executive

# Technical architecture & migration guide
python3 main.py --summarize --days 30 --mode architect

# Changelog digest with export to Markdown
python3 main.py --summarize --days 14 --mode digest --export digest.md

# Ask custom architecture questions
python3 main.py --summarize --query "What are the new capabilities for tabular foundation models and TabFM?"

# Offline analytical mode (no LLM call required)
python3 main.py --summarize --days 30 --offline

---

Project Structure

bigquery-release-summarizer/
├── fetcher.py        # Atom XML feed fetcher, parser & item-level caching
├── summarizer.py     # AI summarizer (via agy CLI) & analytical rule fallback
├── server.py         # Zero-dependency Python HTTP server & REST API
├── main.py           # Unified CLI entrypoint and server runner
├── cache/            # Local JSON cache directory
│   └── release_notes.json
└── web/              # Responsive Web Dashboard
    ├── index.html    # Semantic layout, KPI cards, AI Studio & timeline
    ├── style.css     # Google Cloud dark theme, glassmorphism & typography
    └── app.js        # Dynamic filtering, API integration & Markdown renderer

---

REST API Endpoints

Method	Endpoint	Description
GET	/api/health	Service health status
GET	/api/stats	Aggregate release metrics by category & stage
GET	/api/notes	Query filtered items (days, category, stage, search)
POST	/api/summarize	Generate AI or analytical summary report
POST	/api/refresh	Force invalidation of cache and re-fetch from GCP

---

License

Apache License 2.0. See LICENSE for details.