Problem
- Client context: Private Financial Research Client
- Requested scope: AI Automation & Workflow Automation
- Business tags: AI, Automation, Telegram, Notion, Gemini, Macmini, NewsCollection, Deduplication
포트폴리오
This service is available to members only.
AI-Powered Financial News Collection Automation
Private Financial Research Client
#AI, #Automation, #Telegram, #Notion, #Gemini, #Macmini, #NewsCollection, #Deduplication
Portfolio Overview
This project automates the collection, summarization, classification, duplicate detection, and Notion database archiving of financial news and market updates from Telegram conversations and linked articles.
IT7 built an always-on Mac mini workflow using Telegram API, Gemini API, Notion API, SQLite, and macOS LaunchDaemon. The system allows a non-technical operator to adjust schedule, classification rules, and summary prompts through Markdown configuration files.
Best-fit Use Cases
- Telegram financial news monitoring automation for research teams
- Notion-based news database automation
- AI summarization and duplicate news detection using Gemini API
- Always-on Mac mini automation for internal knowledge workflows
FAQ
Can the system collect from many Telegram channels?
Yes. The collection targets can be refreshed based on the client's Telegram account and operational policy.
Can it summarize messages with multiple article links?
Yes. The workflow extracts linked article content and summarizes each link together with the original Telegram message.
How is the result stored?
The summarized title, original message, summary, URL, classification, status, and duplicate flag are stored as structured rows in Notion databases.
Outcome
The client gained an operational workflow that reduces manual monitoring work and makes financial news searchable by topic inside Notion.
- 2026.06 ~ 2026.07
- Private Financial Research Client