Описание
Accountabilities
- Build and maintain production-grade AI agents using TypeScript, Node.js, AWS Lambda, Step Functions, DynamoDB, Amazon Bedrock, and frameworks such as Strands or LangGraph.
- Design agent workflows by defining what AI models can investigate and which actions must be controlled by deterministic code.
- Develop secure, sandboxed agent runtimes with read-only access, scoped tools, controlled file access, and appropriate budget and timeout limits.
- Evaluate AI features using benchmark datasets, production replays, calibrated evaluation models, and quality gates based on recall, precision, and false-positive rates.
- Compare models and inference effort levels before release to ensure reliable, measurable performance.
- Improve observability through structured logs, reason-coded metrics, alerts, and production troubleshooting using tools such as Splunk and VictoriaMetrics.
- Maintain Terraform infrastructure across environments while respecting tagging standards, permission boundaries, and security requirements.
- Enhance GitLab CI/CD pipelines for building, testing, deploying, and enforcing quality controls.
- Build resilient integrations that handle retries, rate limits, permanent failures, and outdated application state.
- Use AI coding assistants to accelerate development, review generated code critically, and maintain accurate context for AI-assisted workflows.
- Review AI-generated changes for security, correctness, reliability, and maintainability, identif
Полное описание, компания и ссылка на вакансию — после входа.
ВойтиКонтакты работодателя (email/phone/telegram) скрыты из публичного превью.
Отправьте резюме — и работодатели смогут написать вам сами.