Responsibilities (Primary) :

- Build new agent tools/actions using AWS Bedrock Agent Action Groups (Lambda-backed tools).

- Integrate existing internal/external APIs into Action Groups via OpenAPI schemas, standardize contracts (versioning, validation, error shapes).

- Design and implement forecasting input pipelines : ingestion, transformations, feature prep, aggregation, and validation.

- Implement business logic and data transformation layers with strong test coverage.

- Ensure robust error handling, retries, idempotency, and clear observability (logs/metrics/traces).

- Secure tool interfaces (IAM policies, scoped permissions, secrets management) for safe function/tool calling.

- Implement async workflows for long-running/high-throughput work (queues/events).

Essential Skills :

- Python proficiency (must-have) : production services, testing, profiling, packaging, clean architecture.

- Large dataset handling in Python : memory/performance-aware processing (vectorization, chunking/streaming; pandas/NumPy/Polars or similar).

- API development : REST and/or GraphQL, contract-first design, schema validation.

- Database design + query optimization (SQL; Postgres/Aurora/Redshift a plus).

- ETL/ELT patterns focused on correctness, validation, lineage, and maintainability.

- Strong AWS experience : Lambda, API Gateway, IAM, CloudWatch, S3 (Step Functions, EventBridge, SQS/SNS are plus).

- AWS Bedrock Agents concepts : Action Groups, orchestration, tool/function calling, secure interfaces.

- Asynchronous processing with queues (DLQs, retries, idempotency).