ShipStation Global Builds AI IT Help Desk with Amazon Bedrock

Disclosure: this post contains affiliate links. If you buy through them we may earn a commission at no extra cost to you. Learn more.

TL;DR: ShipStation Global built Deckhand, an AI IT help desk agent for Slack using Amazon Bedrock AgentCore. In its first three months, Deckhand resolved 97.5% of channel queries with a 96% positive feedback rating from employees.

What's New

ShipStation Global, the leading intelligent shipping and logistics platform for small and mid-sized shippers, has deployed Deckhand, an AI-powered IT help desk agent built on Amazon Bedrock AgentCore. The solution addresses a common challenge facing IT teams: answering repetitive questions and clearing routine tickets that drain time from more strategic work.

Deckhand operates directly within Slack, where it answers routine IT and company questions from a curated knowledge base, reads error-message screenshots, and creates and updates Jira Service Management tickets. The agent works like any teammate—joining channels, reading conversations, and drawing context from Slack profiles to answer questions in place.

In the 12 months before Deckhand's launch, employees at ShipStation Global posted more than 1,500 threads in the help desk Slack channel. With employees in the US, UK, Spain, Poland, Australia, and the Philippines, more than a fifth of questions arrived outside US business hours. Answers were fragmented across Confluence articles, intranet pages, PDFs, IT-maintained guides, and the channel's own history.

According to Chris Collins, Senior Director IT at ShipStation Global: "We created Deckhand because getting help should be as simple as asking a question in Slack, not deciphering a ticket form or sending an email into the void. We believe that tickets are an indication of operational friction not productivity."

The solution achieved impressive results in its first three months: 97.5% of incoming channel queries were resolved, while earning a 96% positive feedback rating from employees.

Technical Architecture

The solution runs on Amazon Bedrock AgentCore, a platform for building, connecting, and optimizing agents at scale. The architecture consists of four main components:

A Slack service built with FastAPI and Slack Bolt, hosted on AWS App Runner, receives Slack events and builds user-context payloads from requester profiles. An agent service using the Strands Agents framework deploys as a container on Amazon Bedrock AgentCore runtime, managing the multi-turn agent loop using Anthropic's Claude in Amazon Bedrock.

Tools and knowledge are provided through an AWS Lambda function exposed as Model Context Protocol (MCP) tools through AgentCore Gateway, which classifies requests, creates and updates Jira tickets, and analyzes images. An Amazon Bedrock knowledge base indexes content from Confluence, the company intranet, Slack, and marketing pages, storing it in Amazon S3.

Operations and feedback are managed through Amazon DynamoDB for storing feedback, AWS X-Ray for tracing request paths using OpenTelemetry, and Amazon CloudWatch for logging.

Pricing and Availability

Deckhand is a custom solution built by ShipStation Global's engineering team. Pricing information for Amazon Bedrock AgentCore and the underlying AWS services used in the architecture was not disclosed. Organizations interested in building similar solutions can access Amazon Bedrock AgentCore through their AWS accounts.

How to Try It

Deckhand is an internal tool built specifically for ShipStation Global and is not available as a public product. However, the AWS Machine Learning Blog post provides detailed technical documentation on the architecture and implementation. Organizations can use Amazon Bedrock AgentCore to build similar AI-powered help desk agents.

The implementation uses the open source Strands Agents SDK, which allows agents to be built with just a model, a set of tools, and a system prompt. The model ID is set through environment variables, allowing teams to switch models without modifying application code.

Who It's For

This solution is designed for IT teams at companies facing high volumes of repetitive support questions, particularly those with distributed workforces across multiple time zones. Organizations using Slack for internal communication and seeking to improve employee self-service capabilities would benefit from similar implementations.

Companies with knowledge scattered across multiple platforms—such as Confluence, intranet pages, and chat history—can use this approach to consolidate information access into a single conversational interface.

Alternatives

Organizations looking for AI-powered support solutions have several options. Chatbase provides AI chatbot capabilities for customer support. Enterprise teams may also consider building custom solutions with platforms like Make for workflow automation integrated with their existing tools.

Other approaches include traditional IT service management platforms with AI capabilities, though these often require employees to navigate ticket forms rather than conversational interfaces.