Helping Reporters Research and Write at a Regional Newspaper
This supports a fictional regional newspaper that publishes a morning edition in its prefecture, where 60 reporters ask an AI agent by chat and voice for background research before reporting, tabulating statistics and making charts, drafting articles and scheduling interviews. Agents on AgentCore Runtime choose among Gateway tools, Browser, Code Interpreter and external service links, and keep conversation history in Memory. The business scenario is a fictional example prepared by IceShore, not a real company or system. Source: aws-samples/sample-strands-agent-with-agentcore (commit cd66e48). Copyright (c) 2025 Agent Chatbot Template. MIT License (full text in the bundled LICENSE.yaml and at https://github.com/aws-samples/sample-strands-agent-with-agentcore/blob/cd66e4896ee1f65c46bc8c038a8fd89a1dd040ae/LICENSE )
Actors
- Reporters in the newsroom
- IT department staff
- Amazon Bedrock foundation models
- Amazon Nova Act
- External search and data APIs
- Gmail, Google Calendar, GitHub and Notion
- Websites being browsed
Use cases
- Reporters sign in after arriving and open the chat UI with their own settings
- Reporters look up statistics, weather, papers and maps by chat before reporting
- Reporters have the agent collect council schedules and press releases from prefectural and municipal sites
- Reporters hand over fixes to the vote-count tally scripts for election night
- Reporters tabulate population and tourism statistics and make charts for the paper
- Reporters schedule interviews using their own mail and calendar
- Deliver the results of long research into past articles and documents to the reporter's conversation later
- Reporters back from interviews ask questions by voice while sorting their notes
- The IT department builds and deploys containers in the monthly update
- The IT department registers the tools and agents used in the newsroom in the catalog
