AI-Powered Weekly Technology Intelligence Pipeline using Kestra + Ollama + Notion + Slack
OrchestNews is a fully local, zero-cost workflow orchestration project built using Kestra that automatically generates a weekly AI & technology news digest.
The system fetches trending AI/tech news from RSS feeds, filters duplicate articles, summarizes them using a local LLM running via Ollama, extracts macro-level industry insights, ranks articles using lightweight personalization logic, stores everything in Notion, and finally sends a digest notification through Slack.
| Component | Technology |
|---|---|
| Workflow Orchestration | Kestra |
| LLM Runtime | Ollama |
| Model Used | phi3 |
| Storage | JSON + Notion |
| Notifications | Slack |
| Language | Python |
| Deployment | Local + Docker |
| News Sources | RSS feeds |
pip install feedparser requests
notion_upload.py with your Notion API key and database ID.slack_notify.py with your Slack webhook URL.kestra server start
kestra workflow create workflows/weekly_digest.yaml
kestra execution start orchestnews.weekly_digest
docker-compose up -d
docker exec -it orchestnews-kestra-1 kestra workflow create /app/storage/workflows/weekly_digest.yaml
docker exec -it orchestnews-kestra-1 kestra execution start orchestnews.weekly_digest
OrchestNews demonstrates:
It mirrors real-world AI workflow systems used in modern engineering organizations.
I have written a Medium Blog. Do drop in your comments βΊοΈ