-
-
Picture showing background running process
-
Picture showing chat with catch me up on sandbox
Inspiration
Slack is where work happens. Decisions are being made, questions are being answered, and ideas are evolving throughout the day. But after being away for just a few hours—or returning after several days—catching up can feel overwhelming. Important updates can easily get buried among hundreds of messages, making it difficult to understand what actually matters.
I built Catch Me Up to solve this problem. My goal was to create an AI agent that helps people quickly understand the conversations, decisions, and action items they missed without having to manually search through large amounts of message history.
What it does
Catch Me Up is an AI-powered Slack agent that creates personalized summaries of workspace activity.
Instead of reading through every unread message, users can ask the agent to summarize a specific channel or time period, such as:
"Catch me up on #general since yesterday." "What did I miss while I was away?" "Summarize the important discussions from today."
The agent searches relevant Slack conversations, identifies important information, and generates a concise summary highlighting:
Key decisions Important updates Action items Unanswered questions Relevant mentions and ongoing discussions
The goal is to reduce information overload and help users quickly return to productive work.
The goal is to reduce information overload while helping teams stay informed and productive.
How we built it
I built Catch Me Up using Python and Slack's developer platform.
The project uses:
Slack Bolt for Python to handle Slack events and user interactions Slack AI capabilities for intelligent summarization and natural language responses Slack Real-Time Search (RTS) API to retrieve relevant conversations before generating summaries
I designed the application with a modular architecture where search, ranking, summarization, and response generation are separated into different components. This approach makes the system easier to improve, maintain, and extend with future capabilities.
Challenges we ran into
One of the biggest challenges I faced was determining what information is important enough to include in a summary.
A useful catch-up should not simply repeat recent messages. It needs to identify meaningful conversations, highlight decisions, extract action items, and remove unnecessary noise.
Another challenge was balancing summary length and context. The summary needed to be short enough to save users time while still providing enough information to understand what happened.
Accomplishments that we're proud of
I'm proud of building a solution that addresses a real problem faced by teams using Slack every day.
Instead of creating another chatbot that only answers questions, I focused on building an agent that solves a specific workflow problem: helping people recover lost context after being away.
By combining Slack AI capabilities with Real-Time Search, Catch Me Up transforms large amounts of conversation history into concise, actionable summaries.
I'm also proud of creating an experience where users can interact naturally using everyday language instead of complicated commands or manual searches.
What we learned
Building this project gave me practical experience working with Slack's developer platform, event-driven application design, AI-assisted summarization, and search-based information retrieval.
Through this process, I learned that effective AI systems are not just about generating responses—they are about delivering the right information at the right time in a way that helps people make better decisions.
What's next for Catch Me Up
I plan to continue improving Catch Me Up by adding features such as:
Personalized importance ranking Daily and weekly catch-up digests Customizable summary styles Smarter prioritization based on user roles, channels, and activity Integration with external knowledge sources through MCP Improved context awareness across conversations
My vision is to make Catch Me Up a reliable AI teammate that helps people stay informed without spending hours searching through messages.
Log in or sign up for Devpost to join the conversation.