Narasimhan Venkadeswaran

Site Reliability Engineer

Domino Data Lab, San Francisco, CA 2021 – Present

• Manage and optimize infrastructure for Domino AI, an enterprise MLOps platform, ensuring high reliability and performance.
• Develop monitoring solutions for critical AI applications, leveraging tools like Prometheus, Grafana, and the ELK stack.
• Automate deployments and configuration management using Terraform and Ansible, reducing manual effort and errors.
• Design CI/CD pipelines for continuous training and deployment of machine learning models using Jenkins and Argo CD.
• Implement cost-saving strategies on Google Cloud for Vertex AI by automating resource optimization for idle models and data sources.
• Serve as Embedded SRE within multiple teams, including Infrastructure, IdSM, and Domino Cloud, enhancing collaboration and system efficiency.
• Maintain security compliance by implementing best practices in networking and data protection.

• Led Kubernetes-based API middle-tier systems, serving 50,000 RPS with 99.99% uptime.
• Implemented canary deployments and automated ingress routing for robust, performant deployments.
• Designed and scaled a 30TB Cassandra backend across five data centers, achieving sub-500ms latency and supporting 100,000 RPS for Yahoo's Media Content Platform.
• Spearheaded Big Data infrastructure automation with Terraform and Kubernetes, optimizing batch and real-time analytics pipelines for 5+ million daily events.
• Conducted log analysis with Splunk and ELK, reducing incident resolution time by 40%.
• Collaborated across teams to implement A/B testing frameworks for traffic delivery and large-scale deployment paradigms.
• Improved security and compliance by automating CVE detection and integrating OWASP best practices into CI/CD pipelines.