AWS Professional Services built four migration agents; infrastructure code drops from 3-4 weeks to minutes
AWS consultants describe four agents for cloud migrations: intake, infrastructure code, governance and post-cutover monitoring. Infrastructure code per app falls from 3-4 weeks to minutes. Humans approve actions.
- Organization
- Amazon Web Services (Professional Services)
- Industry
- Cloud consulting
- When
- Oct 2026
- Evidence
- Vendor case study
Reported outcome
Infrastructure-as-code development per application fell from 3 to 4 weeks to minutes, from AWS's internal project tracking data, across a portfolio of over 300 applications.
- 3-4 weeks to minutes (company)IaC development per application
- over 300 applications (company)Portfolio reported on
Caveats
Published by the company that sells the agent. Treat numbers as marketing until confirmed elsewhere.
- AWS wrote this and sells the platform. The timing data is 'internal project tracking data' with no method given.
- The post says 'Your results might vary based on portfolio size and tool integrations.'
- 'Minutes' is for code generation. Reviews and approvals still take time. The post says no agent acts on its own on production systems.
- The code samples are illustrative, not production code.
What the agents do
An Intake agent reads architecture documents and questionnaires and drafts a target design with sizing and a compliance report. An IaC agent writes infrastructure code from the organization's approved modules, checks each action against policy, and pushes it to a repository for normal review. A governance agent collects status from Jira, Confluence and Webex. An SRE agent watches workloads after cutover and suggests fixes.
What happened
The pattern runs next to AWS Transform and AWS DMS. Write actions need human approval. The post says to check first that a managed AWS service does not already cover your migration path.
What to copy
- Build agents from your approved modules, not free-form code.
- Check each action against policy before it runs.
- Run a pilot wave to learn the cost per application first.
Starter prompt
Written by us, not by AWS.
Application description: {app_summary}
Approved infrastructure modules: {module_list}
Draft the infrastructure plan using only these modules.
List any need that no module covers. Do not invent a module.
Mark every step that needs human approval.
Sources
- Scaling cloud migrations with agentic AI on Amazon Bedrock AgentCore — AWS Machine Learning Blog, Oct 1, 2026
Added Oct 26, 2026. Spotted an error? Send a correction.
Get new use cases weekly
One short email. Sources included.
Related use cases
Coupa's Sentinel runs 234 engineering agents; 75% of support queries resolved without a person
500+ users, 234 agents and 15,600+ automated tasks a week; 50% less ticket investigation time and 75% of support queries resolved autonomously.
Reactiv's three-agent AI Scheduler cut merchant app setup time by 80%
Reactiv reports an 80% cut in merchant configuration time, 33% faster time to production, and scheduled jobs dropping from over 10 minutes to about 5.
Formula 1's agents cut data-source onboarding from 6-8 weeks to about 40 minutes of code generation
Data-source onboarding fell from 6-8 weeks to about 40 minutes of code generation plus hours of review; the post says agents do 95% of onboarding tasks unaided.