Harness announces new AWS integrations to connect AI code creation with intelligent delivery
AI is speeding up how code gets written. The bottleneck is getting that code safely into production. Testing, security checks, and deployments still consume cycles and rely on manual steps.
Harness and AWS are closing that gap with integrations that tie AI-assisted development directly to delivery workflows. The result: a continuous loop from code to deployment that learns, adapts, and improves over time.
What's new
- Kiro + MCP integration: Harness's Model Context Protocol (MCP) Server connects with Kiro so developers can manage CI/CD, investigate security issues, list and run chaos experiments, and analyze cloud costs using natural language directly from AWS environments and their IDE/CLI. Learn more about MCP on the official site.
- Harness SaaS on AWS: Run the platform on AWS with unified governance, security, and performance. Teams can streamline procurement via AWS Marketplace.
- Delivery intelligence with a Knowledge Graph: Ongoing work embeds developer checks, security, and automation earlier in the lifecycle, making delivery more adaptive and context-aware with AWS data.
Why it matters for engineering leaders
More AI-generated code means more changes to test, secure, and ship. These integrations reduce tool switching, bring delivery context to where devs work, and automate the grind in pipelines.
Platform teams gain consistent governance across AWS environments, while improving traceability from code to production. Procurement and rollout also get simpler with Marketplace.
How the Kiro + MCP workflow plays out
- Ask, "Why did the build fail?" and get the failing step, logs, and suggested fixes.
- Trigger a safe retry, rollback, or a pre-approved remediation directly from your IDE or CLI.
- Query "Show open security issues for Service X" and get prioritized findings with links to pipelines and commits.
- List and run chaos experiments, then capture outcomes as feedback into delivery checks.
- Analyze cloud costs by service, environment, or recent deployments to spot regressions tied to code changes.
Running the platform on AWS
Enterprises can operate the platform fully within their AWS footprint with centralized visibility and policy control. Alignment across compute, storage, and observability keeps performance predictable and governance consistent.
For teams already building on AWS, this creates a direct path from AI-assisted development to production delivery without bolting on separate systems.
Roadmap highlights
- A rules-and-learning system that detects recurring pipeline errors and prevents repeats automatically.
- Shift-left checks that bring policy, security, and workflow automation into the code generation process.
- Expansion of the Software Delivery Knowledge Graph with AWS data for stronger context across services, environments, and deployments.
What leadership says
"The core mission of Harness is to simplify software delivery so teams can move faster with confidence," said Bala Venkatrao, Senior Vice President and General Manager of Platform. "Our ongoing collaboration with AWS takes that vision even further. By collaborating with one of the most trusted ecosystems in software development, we're delivering a unified experience that connects AI-powered software development with intelligent delivery, enabling customers to deploy software innovation into production faster, safer, and more efficiently than ever before."
How to get started
- See the integrations live at AWS re:Invent 2025 (Dec 1-4).
- Join the joint webinar on December 16, 2025 at 1:00pm EST.
- Procure via AWS Marketplace and connect MCP-enabled workflows with Kiro. For MCP details, visit the Model Context Protocol site.
Related resources
Building your team's AI delivery skills? Explore hands-on resources for AI-assisted coding and tooling on Complete AI Training.
Your membership also unlocks:
