How ShipStation Global built an AI IT help desk on Amazon Bedrock AgentCore
ShipStation Global's engineering team built Deckhand, an AI IT help desk agent for Slack on Amazon Bedrock AgentCore. It answers routine IT questions from a curated knowledge base, reads error-message screenshots, and files Jira tickets.
ShipStation Global developed Deckhand, an AI-powered IT help desk agent for Slack, built on Amazon Bedrock AgentCore. It automates routine IT tasks, answers questions, and creates Jira tickets, improving efficiency for IT teams and employees.
What happened
ShipStation Global's IT team developed Deckhand, an AI-powered help desk agent for Slack, to automate routine IT tasks. The agent uses Amazon Bedrock AgentCore to answer questions, analyze screenshots, and create Jira tickets, significantly reducing the workload on IT teams.
Why it matters
Deckhand improves IT efficiency by automating repetitive tasks, allowing IT teams to focus on more complex issues. It enhances employee experience by providing immediate assistance and reduces the time spent on ticket management.
- Leverage AI to automate routine IT tasks and improve efficiency.
- Use managed services like Amazon Bedrock AgentCore to reduce infrastructure complexity.
- Integrate multiple tools and knowledge sources for a comprehensive solution.
- Continuous feedback and iteration improve AI system performance and user satisfaction.
- AI agents can significantly reduce operational overhead and enhance user experience.
The bigger picture
Deckhand reduced IT ticket volume by 812 conversations, saving approximately 1,600 hours annually. It improved employee satisfaction with a 96% positive feedback rating and streamlined IT operations by automating routine tasks.
ShipStation Global's Deckhand demonstrates how AI can streamline IT operations by automating routine tasks, improving employee satisfaction, and reducing workload on IT teams. The solution leverages Amazon Bedrock AgentCore and other AWS services to create an efficient, scalable help desk agent.
Technical details
- Amazon Bedrock AgentCore runtime hosts the agent in isolated sessions without servers or containers.
- Strands Agents SDK is used to build the agent with a model, tools, and system prompt.
- Amazon Bedrock Knowledge Bases indexes content from Confluence, intranet, Slack, and marketing pages.
- AWS Lambda functions handle classification, ticket creation, and image analysis via AgentCore Gateway.
- Recursive Language Models (RLMs) mine Slack history to extract structured data for the knowledge base.
FAQ
What is Deckhand?
Deckhand is an AI-powered IT help desk agent for Slack developed by ShipStation Global, built on Amazon Bedrock AgentCore. It automates routine IT tasks, answers questions, and creates Jira tickets.
How does Deckhand improve IT efficiency?
Deckhand reduces the volume of IT tickets by handling 97.5% of incoming queries, saving approximately 1,600 hours annually. It automates repetitive tasks, allowing IT teams to focus on more complex issues.
What technologies does Deckhand use?
Deckhand uses Amazon Bedrock AgentCore, Strands Agents SDK, Amazon Bedrock Knowledge Bases, AWS Lambda, and Recursive Language Models (RLMs) to mine Slack history and extract structured data.
How does Deckhand handle feedback?
Deckhand includes feedback buttons for users to rate responses. Negative feedback is used to improve the system prompt and correct knowledge base information.
What are the key benefits of Deckhand?
Key benefits include improved employee satisfaction with a 96% positive feedback rating, reduced IT workload, automated ticket creation, and efficient use of managed AWS services.
This summary and analysis were generated with AI from the original article and may contain errors. Read the original for the full story.
Read the full story
Originally published by AWS ML Blog on Oct 13, 2026.
Comments
Sign in to join the discussion.