LT
Job Description
- RCA Agent Development
- Design, develop, and enhance the RCA agent using GenAI agentic AI patterns
- Build prompts, reasoning workflows, tool/function integrations, and supporting APIs
- Optimize agent accuracy, reliability, latency, safety, and cost
- AD RCA Integration
- Integrate the upstream AD workflow with the RCA agent through enterprise APIs/service layers
- Implement data and workflow integration between the agent and dependent enterprise systems
- Validate end-to-end integration and error-handling scenarios
- Evaluation Framework
- Build an automated evaluation framework and test harness for the RCA solution
- Define acceptance criteria and metrics covering accuracy, factuality, grounding, safety, latency, and cost
- Perform regression, adversarial, hallucination guardrail, and failure-path testing
- Use evaluation findings as release gates and inputs for system improvement
- E2E Tuning & Optimization
- Analyze failures across the complete RCA pipeline and tune prompts, context, orchestration, and inference behavior
- Improve accuracy and production reliability based on evaluation results
- Identify error propagation across multi-step agent workflows through tracing and observability
- AWS SageMaker & Bedrock
- Develop and operationalize agent workloads using Amazon Bedrock where applicable
- Support model deployment optimization, monitoring, and lifecycle activities using SageMaker
- Optimize prompts, embeddings, and inference performance
- Establish monitoring for production performance and anomalies
- Short version for JD RR
Key Responsibilities
- Develop and enhance the RCA agent, integrate AD-to-RCA workflows, build the GenAI evaluation framework, perform end-to-end testing and tuning, optimize prompts/model inference, and support deployment monitoring and operationalization using AWS SageMaker and Amazon Bedrock
Mandatory Skills : AI platform engineering
Sponsored