Amazon Q Developer can support planning, coding, testing, documentation, review, deployment, and debugging across the software development life cycle.
2
The demo builds a 2048 game with Q Developer in the CLI and IDE, then adds tests, documentation, deployment scripts, and bug fixes.
3
AI can handle repetitive development work, but engineers still need to review generated code, plan infrastructure, and use Q responsibly.
Summary
Vikash Agrawal and a developer advocate from AWS demonstrate how Amazon Q Developer fits into the software development life cycle. They begin with a 2048 game built in Python with FastAPI and Poetry through the CLI. Q creates a plan, generates the project, moves it into the requested folder, and writes a README. In the IDE, /test generates unit tests, /dev fixes a reported bug, and /doc creates updated documentation with API examples and a data flow diagram. The demonstration then moves to GitHub, where Q can work from issues, open pull requests, respond to review comments, and run code scans. For deployment, Q generates AWS SAM scripts. When the deployed application produces an import error in Lambda, Q Developer in the AWS console examines CloudWatch logs and associated resources to identify the cause. The speakers are honest that planning, infrastructure design, code review, prompt design, and responsible use remain the engineer's responsibility.
Amazon Q Developer can be used throughout the software development life cycle
The speakers frame software development as a cycle that begins with planning, moves through creation and testing, and continues into deployment, maintenance, and modernization. Amazon Q Developer is presented as a coding assistant that works in an IDE, a CLI or terminal, and GitHub. They also show it in AWS operations tooling. The point is to use one assistant across separate stages rather than limit it to code completion. The speakers describe planning as requirements gathering and product feedback, while testing includes end-to-end, integration, and penetration testing.
The CLI can plan and build a 2048 game from a natural-language request
The demo starts Q in the command line with the command q. Users can chat with it in natural language or inspect available commands with /help. The speakers ask it to build a 2048 game in Python using FastAPI and Poetry. Q produces a plan that can be accepted one step at a time or trusted to run. During the demo, it creates the project structure, API code, response models, POST APIs, and a README. It can also move the generated project into a development folder after the user realizes it was created in the wrong location.
Generated code still needs review before production use
After Q creates the initial game, the speakers ask whether the result is production ready and answer no. They recommend reviewing generated code in a real production setting, even though they trust it for the demonstration. The demo then opens the project in an IDE, where Q can work with the existing folder structure and project context. The speakers also mention that Amazon Q Developer supports the IDE preferences used by their engineering teams, including VS Code, IntelliJ, and Eclipse.
The IDE agents add tests, fix bugs, and update documentation
In the IDE, /test invokes the Amazon Q Developer agent to inspect the project and generate unit tests. The speakers say it understands the code's state management and creates tests for that behavior. They then describe /dev as a feature-development command that uses the project as context. A short request containing an error message leads Q to locate the bug and propose a code change, which the presenter applies. Finally, /doc generates a README from the source and tests. The resulting documentation includes API examples, troubleshooting information, and a data flow diagram for state management.
Amazon Q Developer can also work without leaving GitHub. The speakers show an installation from the Q Developer marketplace, followed by repository selection. A user can connect Q to an existing issue or create a new one, apply a development-agent label, and wait for a pull request. The resulting PR shows changed files. Reviewers can comment and request changes, after which Q performs another update. The speakers also say Q automatically runs code scans and fixes issues directly in the pull request.
Deployment planning belongs in the development workflow
For deployment, the demo uses the /dev agent in the IDE to generate AWS SAM scripts and deploy the application into AWS. The speakers connect this step to the familiar problem of software that works on a developer's machine but fails after deployment. They conclude that infrastructure should be considered earlier, rather than added only after the application has been written. Vikash Agrawal says the planning phase could include asking Q for a production-ready approach and iterating on that plan before generating and saving the code.
Q Developer can investigate a Lambda failure through CloudWatch
After deployment, the application does not run correctly in production. In the AWS console, the presenter opens CloudWatch, checks the Lambda logs, and finds an import error. Amazon Q Developer's AI operations feature examines the log group and associated resources, creates a topology, and identifies a root cause. The speakers describe this as a way to understand an incident without manually searching for the right timestamp or preparing a separate incident report. The result gives the developer information to take back into the application and fix the deployment.
Engineers remain responsible for plans, prompts, and safe use
The closing discussion puts limits around the automation shown in the demo. The speakers say engineers and managers still plan the work, while AI helps with tedious tasks. They recommend planning cloud infrastructure early, using security scans and monitoring, and iterating before code is generated. They also say prompt engineering may become an important skill because prompt design affects the output from models and agents. Q Developer is presented as free to try in the IDE, CLI, and GitHub, but the responsibility for desired and responsible results remains with the person using it.
"Amazon Q developer is a coding assistant that is also in your IDE, can be even in your CLI and terminal, actually also has a GitHub extension and can be also in different parts of the life cycle."01:17
Who should watch
You want to try an AI coding assistant across the CLI, IDE, GitHub, and AWS console rather than use it only for autocomplete.
Your team is considering generated tests, documentation, pull requests, deployment scripts, or operational debugging and needs to see a complete example.
You need a reminder that generated code still requires review and that infrastructure, prompts, monitoring, and security belong in the plan.
