Dharmender Singh

Built a secure log-export and audit pipeline for LLM systems using AWS Lambda and AWS Step Functions, exporting logs to Amazon S3 with encryption and access controls to improve traceability, security, and compliance, while leveraging dynamic wait states for asynchronous reliability. Implemented CloudWatch Logs and CloudWatch Metrics for end-to-end observability of Lambda/Step Functions, with alarms and dashboards for real-time performance monitoring. Developed an agentic SDLC automation framework using Enterprise Claude (Anthropic) with custom plugins and Jira/Azure DevOps API integrations to automate command execution and lifecycle tracking across pipelines. Deployed Amazon Bedrock AgentCore solutions (MCP servers) to orchestrate multiple AI agents at scale, enabling centralized management and context sharing. Created a centralized monitoring dashboard to track token usage, cost trends, and operational tickets across all LLMs behind Bedrock for leadership insights.

Developed Pytest-based integration and regression test suites for networking devices and infrastructure, ensuring reliability and compliance through automated CI/CD testing. Contributed to CI/CD pipelines and AWS CDK for automated infrastructure deployment, and designed a dashboard for AWS account spend visibility to enable cost oversight. Participated in hiring processes and authored migration guides to optimize resource allocation and minimize service disruption.

Implemented and managed VM infrastructure across Azure and AWS, automating configuration with PowerShell, managing IAM, Key Vault, AD, storage, monitoring, backups, Site Recovery, and Log Analytics/Application Insights. Automated system administration tasks with Bash, Python, and Azure CLI; provisioned on-demand cloud environments using Terraform (VPCs, Subnets, SGs, etc.). Used Ansible and scripting for configuration management and CI/CD automation; containerized applications with Docker and orchestrated with Kubernetes; implemented monitoring with Azure Monitor, Prometheus, and Grafana; adhered to IaC with Terraform and basic network security practices.