Comma: Smart AI Tech Curation

Comma: Never a Period, Always a Comma

In the midst of an overwhelming volume of newsletters and countless tech sources, staying updated has become a challenge. Comma is a serverless curation bot designed to solve this by delivering a single, carefully selected article about AI and technology daily, ensuring relevance without the noise.

The name reflects the project’s core philosophy: in writing, a comma signifies a pause, not a full stop. It suggests that the journey of knowledge is continuous—there is always another idea to explore, another connection to be made.

The Challenge & Evolution

The project originated from my personal difficulty in keeping up with high-quality tech news in an industry that evolves daily.

An initial version of the bot failed due to low adoption, which provided a valuable lesson: Technical excellence isn’t enough without a great User Experience (UX). This led to a pivot focusing on a seamless, cross-platform experience that integrates directly into the user’s daily workflow via messaging platforms.

Technical Architecture

The system follows a serverless logic designed for efficiency, reliability, and high-quality content synthesis.

graph LR
    A[Timer Trigger] --> B[Python Logic / FastAPI]
    B --> C{Database Check}
    C -- New Article --> D[Groq LLM / CoD Prompt]
    C -- Duplicate --> B
    D --> E[Langsmith Observability]
    D --> F[Telegram API]
    
    subgraph Azure Cloud
    B
    C
    end
    
    subgraph AI Layer
    D
    E
    end

To ensure the summaries are both concise and informative, I implemented a GenAI/LLM (Groq) pipeline using the Chain-of-Density (CoD) prompt strategy. This state-of-the-art approach allows the bot to condense content progressively without losing critical entities or facts.

Key Features:

• Serverless Execution: Operates via a Timer Trigger for fully automated daily delivery.
• Smart Selection: Uses random selection from curated sources with “Retry loop” logic to guarantee a fresh article every day.
• Duplicate Prevention: Integrated database connection to ensure the same article is never sent twice.
• Cross-Platform Delivery: Designed to send messages across multiple platforms, currently integrated with Telegram.

Tech Stack

• Language: Python
• Frameworks: FastAPI
• AI/LLM: Groq (using Chain-of-Density prompting)
• Infrastructure: Docker & Uvicorn
• Cloud: Azure
• Observability: Langsmith

Future Roadmap

The next phase for Comma involves implementing a full CI/CD pipeline with GitHub Actions to automate deployments and further refining the recommendation engine based on user engagement.

Link to GitHub Repository

Tech Stack: Python Azure Docker FastAPI Groq