Developer (IN)
HCLTech
HCLTech
HCLTech is seeking a skilled AI Builder to develop an advanced AWS DevOps Agent. This agent will leverage Generative AI to automate root-cause analysis (RCA) and other critical DevOps operations across AWS environments.
Working within the GenAI delivery build layer, you will transform approved architectures into functional, tested, and production-ready agent capabilities. This involves creating reasoning chains for RCA, correlating logs and metrics, suggesting remediation steps, assessing deployment risks, and summarizing incidents using AWS-native AI and operations services integrated with LLM orchestration.
Collaboration is key in this role. You will partner closely with the Enterprise AI Architect and DevOps/SRE teams to deliver reliable, trustworthy automation solutions swiftly and maintain a high standard of quality.
Develop and implement the core RCA capability of the AWS DevOps Agent, ingesting logs, metrics, traces, and deployment events to pinpoint incident root causes.
Create agentic workflows and tool-calling integrations, enabling the agent to interact with AWS services for diagnostics and remediation.
Expand the agent's functionality beyond RCA to include anomaly detection, deployment risk scoring, incident summarization, and automated runbook execution.
Integrate the agent with AWS operational tooling like Amazon DevOps Guru and AWS Systems Manager Automation, adhering to architectural specifications.
Craft and refine prompts, function definitions, and retrieval pipelines to ensure the agent delivers accurate and explainable RCA narratives within specified latency and cost parameters.
Implement robust guardrails and human-in-the-loop approval processes to ensure safe and autonomous changes to production environments.
Write comprehensive unit, integration, and evaluation tests to monitor and maintain the accuracy of RCA and other quality metrics.
Support production readiness by contributing to the agent's observability, deployment scripts, and rollback procedures.
Troubleshoot and resolve issues within existing agent capabilities, addressing reasoning regressions and integration problems.
Document agent capabilities, prompt/tool versions, and configurations to ensure maintainability and auditability.
Collaborate with stakeholders on feasibility assessments, effort estimations, and technical trade-offs during design phases.
Leverage and extend approved architectural patterns for agentic DevOps automation.
Possess 3-7 years of experience in software or AI engineering, with a strong track record in building Generative AI or agentic applications, preferably within DevOps, SRE, or observability contexts.
Demonstrate practical expertise in agent frameworks, tool/function calling (e.g., Amazon Bedrock Agents, LangChain/LangGraph), and prompt engineering for reasoning tasks like RCA.
Have working knowledge of AWS observability and operations services, including CloudWatch, X-Ray, CloudTrail, AWS Config, and Amazon DevOps Guru.
Exhibit strong Python skills and proficiency in working with AWS SDK (boto3), REST APIs, and orchestration frameworks.
Possess experience in correlating logs, metrics, and traces for incident diagnosis, understanding common failure modes in distributed cloud systems.
Understand fundamental security and safe-automation practices for agentic systems, including action approval gates and least-privilege execution.
Be familiar with testing and evaluation approaches for GenAI and agent outputs, such as automated evaluations and human-in-the-loop reviews.
Have working knowledge of CI/CD on AWS and basic infrastructure-as-code concepts.
Ability to interpret architecture and design specifications, ask pertinent questions, and identify feasibility issues promptly.
Excellent written communication skills for documentation and team handoffs.
Experience with incident management/RCA tooling or exposure to multi-agent orchestration is beneficial. Familiarity with LLMOps/MLOps tooling and basic Responsible AI concepts is also advantageous.
HCLTech
Information Technology & Services