The Future of Automated Journalism.
Monitor wires, discover trends, and synthesize high-fidelity reports using a professional AI-native editorial surface.
What It Does • Tech Stack • Getting Started • Editorial Logic • Connect
AutoPress is a fully automated AI-native news surface. It monitors global wire services and performs state-of-the-art synthesis to deliver high-authority reporting without human intervention.
| Output | Description |
|---|---|
| Trend Discovery | Real-time monitoring for HOT, RISING, or OUTLIER news |
| In-Depth Reporting | 700-word synthesized reports with historical context & stakeholders |
| Viral Snapshots | Automated 15-second high-impact social media briefings |
| Newsroom Intel | Sidebars explaining the "Why" behind momentum and velocity |
| Smart Q&A | Context-aware sidebar for deeper interrogation of AI reports |
| Daily Digest | A clean, high-velocity morning briefing for rapid coverage |
- Editorial Light (Ink & Cream): A premium "Newsprint" aesthetic with high-contrast typography designed for trust and professional readability.
- Proactive Background Pipeline: High-performance architecture that authors deep reports asynchronously in the background. The homepage loads instantly with skeletons, polling the pipeline without blocking.
- Newsprint Loading Skeletons: High-fidelity shimmer animations maintain visual interest and visual hierarchy during real-time data discovery.
- Discovery Intelligence Hub: Interactive dashboard for monitoring story velocity and impact across various categories.
- Mobile-Authority Design: Responsive vanilla CSS that guarantees sharp, high-contrast reading layout on every device.
| Layer | Technology |
|---|---|
| Framework | Next.js 14 (App Router) |
| Editorial AI | Groq API – llama-3.1-8b-instant (Two-Pass Generation) |
| News Data | NewsAPI – International Wire Service |
| Persistence | File-based cache (.cache/articles.json) |
| Styling | Vanilla CSS (Editorial-Light System) |
| Icons | Lucide / Custom SVG Branding |
AutoPress/
├── src/app/ # Next.js App Router (Pages & Routes)
│ ├── article/[slug]/ # Client-side article viewer (SWR)
│ ├── digest/ # Morning briefing portal
│ └── layout.tsx # Global branding & SEO metadata
├── src/components/ # Reusable UI (Skeletons, Cards, Modals)
├── src/lib/ # Core logic (Groq rotation, Newsroom Pipeline)
├── docs/ # Architecture documentation
│ ├── HLD.md # High Level Design
│ └── LLD.md # Low Level Design
├── .cache/ # Local persistence for articles & highlights
├── .env.local # Private API keys (Groq, NewsAPI)
├── LICENSE
└── README.md
| Document | Description |
|---|---|
| High Level Design (HLD) | System architecture, pipeline flow, and AI editorial strategy |
| Low Level Design (LLD) | Pipeline background worker, Two-pass LLM synthesis logic, and SWR data schema |
- Node.js 18+
- NewsAPI Key
- Groq API Key
git clone https://github.com/TuShArBhArDwA/AutoPress.git
cd AutoPress
npm installCreate a .env.local file in the root directory:
NEWS_API_KEY=your_news_api_key_here
GROQ_API_KEY=your_primary_groq_keynpm run devVisit http://localhost:3000
- Discovery Pass: Engine monitors wires for Story Density. Articles exceeding the momentum mean are marked as Rising.
- Authorship Pass: Two-Pass generation via
llama-3.1-8b-instantauthors 700-word reports. Pass 1 extracts JSON metadata; Pass 2 writes the narrative. - Verification: Reports are assessed for factual grounding and source cross-referencing.
- Synthesis: Final output includes Social briefings, Timelines, and Impact scores.
- Token Efficiency: Tiered generation (Highlights vs. Full Reports) saves 90% of token budgets.
- Isolated Pipeline: Full architectural separation between data gathering and creative authorship.
This project is licensed under the MIT License - see the LICENSE file for details.
If you’d like to connect or provide feedback, feel free to reach out — Click here