Build a coding agent that fixes bugs using Agno, xpander.ai, and Amazon Bedrock

​Frameworks like Agno make it easy to build autonomous AI agents. But shipping a production-ready agent that integrates with your tools, writes usable code, and scales across your environment still takes too much work.

​In this workshop, you’ll build and deploy a Slack-native coding agent that monitors bug reports (from GitHub Issues or Notion), proposes a fix, and shares it with your team—using the Agno framework, deployed with xpander.ai, and powered by SOTA Models from Amazon Bedrock / OpenAI or Anthropic.

​What You’ll Build

​An agent that can:

• ​Watch GitHub Issues or a Notion bug board for new reports

• ​Parse and understand bug descriptions

• ​Generate and test a code patch using Bedrock models or Codex as a tool

• ​Post the proposed fix in Slack for review

• ​Run in the cloud with secure, observable infrastructure

​Agenda

​Session 1: Demo (30 min)

​How production-grade coding agents work

​We’ll walk through a live agent that watches for bugs and produces patches in response—showing how Agno handles logic, how Bedrock powers generation, and how xpander.ai takes care of runtime, memory, and Slack integration.

​Session 2: Hands-On (2 hours)

​Build your own bug-fixing agent

​You’ll start from a pre-configured Agno template and add:

• ​GitHub or Notion integration (your choice)

• ​Slack delivery and user auth

• ​Custom tool logic and cloud deployment

​By the end, your agent will be running in Slack, ready to spot a bug and propose a fix.

​Who Should Attend

• ​Developers building custom AI tools

• ​DevEx teams exploring Slack-native workflows

• ​Engineers looking to automate repetitive debugging or patching tasks

​Requirements

• ​Laptop

• ​Some coding experience (We can make you ship agents like the best AI developer but it's not a no-code workshop!) 

• ​GitHub account and/or Notion board with test issues

• ​Slack workspace access

• ​Physical government-issued ID (venue requirement)