Slack Context Agent

Architecture Diagram
Project Overview

Inspiration

Teams rely heavily on Slack for collaboration, but important decisions and action items often get buried in long conversations. I wanted to build an AI assistant that could understand the context of a discussion and provide quick summaries and answers without requiring users to scroll through the entire chat history.

What it does

Slack Context Agent retrieves recent messages from a Slack channel, builds conversation context, and uses a Large Language Model (LLM) to answer user questions.

The agent can:

Summarize discussions
Extract action items
Identify pending tasks
Answer context-based questions such as "Who is responsible for deployment?"

How I built it

The application was built using Java and Spring Boot.

The Slack Bolt SDK listens for app mentions and retrieves recent channel messages through the Slack API. These messages are processed into a structured prompt, which is sent to an LLM through OpenRouter. The generated response is then posted back into the Slack conversation.

Challenges I ran into

One of the biggest challenges was integrating multiple services while ensuring reliable context retrieval. Another challenge was handling external AI provider limitations such as API rate limits and making sure the application handled these scenarios gracefully.

Accomplishments that I'm proud of

Built an end-to-end Slack AI assistant in Java.
Successfully integrated Slack APIs with an LLM.
Implemented context-aware summarization and question answering.
Designed a modular architecture separating Slack integration, context retrieval, prompt generation, and AI communication.

What I learned

Through this project I gained hands-on experience with Slack app development, event-driven programming using the Slack Bolt SDK, prompt engineering, REST API integration, and building AI-powered applications using Spring Boot.