Sriram Maheswaran
San Ramon,CA
Summary
Principal Engineer with 18+ years in software engineering and 8+ years at Apple, operating Kubernetes infrastructure at fleet scale (2,000+ clusters), architecting AI/ML agent platforms, and rolling out end-to-end observability and distributed tracing. Hands-on technical leader with deep expertise in deployment automation, cluster lifecycle, and OpenTelemetry/Jaeger tracing—mentors engineering teams and drives architecture across organizations.
Overview
21
21
years of professional experience
Work History
Principal Engineer
Apple Inc
12.2017 - Current
- Operate Kubernetes infrastructure at fleet scale—2,000+ clusters across multiple regions supporting engineering, AI/ML, and external-facing user workloads. Own cluster lifecycle automation (provisioning, upgrades, decommissioning), capacity planning, and multi-region HA.
- Led the original design, build-out, and migration of legacy workloads to Kubernetes-based platforms—modernizing the deployment model and improving reliability and scalability fleet-wide.
- Hardened cluster operations across the 2k+ cluster footprint—HA, fleet-wide rolling upgrades, capacity planning, GitOps-based config management, and incident response—for production-critical services.
- Manage external-facing Apple web properties on Kubernetes at global scale—owning scalability, high availability, and reliability across high-traffic user-facing workloads.
- Run the Nginx-based ingress and traffic-handling layer fronting these services for routing, TLS termination, and resilience.
- Operate Spinnaker-based CI/CD pipelines and Kubernetes deployments enabling rapid, safe rollouts—automated promotion across environments, canary and blue-green strategies, and automated rollback; led CI/CD automation across multiple internal services that cut manual release effort and accelerated delivery cycles.
- Standardized container-based build and test agents, enabling reproducible builds and consistent environments from dev through production.
- Led the monitoring and observability function—defined alerting standards, designed dashboards, and rolled out telemetry across services using Prometheus, Alertmanager, and Grafana.
- Designed and rolled out end-to-end distributed tracing with OpenTelemetry (OTEL) instrumentation and Jaeger—giving engineering teams cross-service request traces, latency breakdowns by hop, and dramatically faster root-cause identification across multi-service workflows.
- Established SLO/SLI-based reliability practices, reducing alert noise and improving mean-time-to-detection and recovery across the fleet.
- Use Catchpoint synthetic monitoring to validate availability, performance, and end-user experience from diverse global vantage points—proactive detection of regressions before customers are impacted.
- Architect and own Apple's company-wide AI/ML agent platform—the foundation for agent-based applications across the organization.
- Provide technical architecture mentorship to a team of 5 engineers: review system designs, guide architectural decisions, and partner on coding and design improvements.
- Designed the reusable agent framework on Apple's AI/ML platform—intent-based routing dispatches requests to specialist agents via declared capability metadata, automated skill/capability discovery makes new agents routable without manual wiring, and a declarative config-driven onboarding model lets teams add agents by changing config only (no platform code changes).
- Integrated mem0 for persistent cross-session agent memory; adopted LangGraph for graph-based multi-step orchestration (stateful execution, conditional branching, retries); standardized on the A2A (Agent-to-Agent) protocol for inter-agent communication, enabling multi-agent workflows that span team and organizational boundaries.
- Built Claude Agent SDK-powered agents for SRE automation—on-call alert analysis correlates signals across systems and ranks root-cause hypotheses (cutting time-to-triage), plus knowledge-graph dependency reasoning over services and infrastructure for impact analysis, change-risk assessment, and faster incident response.
- Architected and built Apple's internal RAG platform from scratch on Milvus—chunking strategy, pluggable ingestion/retrieval architecture (custom embedders, retrievers, post-processors), and ANN/sharding tuning for production-scale vector search. Own reliability and onboarding; adopted across multiple internal organizations.
- Manage NVIDIA Triton Inference Server deployments serving the Apple Maps organization—production model-serving infrastructure for Maps' ML/AI workloads.
- Led cross-functional teams to drive product development and enhance system architecture.
Staff Software Engineer
Juniper Networks
02.2013 - 12.2017
- Designed and maintained deployment infrastructure (SaltStack/Pillar/Jinja) across the Cloud Service-Platform product; managed Docker, Kubernetes, Keystone, RabbitMQ (cluster + federation), Cassandra, and Swift.
- Led an AWS provisioning project using Terraform—owned design and team mentoring; delivered in a month and received the Business AMP company-level award.
- Implemented HA + scalability (HAProxy/Keepalived, K8s master, SaltMaster, ETCD); cut deployment time from 2 hours to 40 minutes via parallel logic that avoided HA race conditions.
- Introduced Reprepro (local Debian repo) to eliminate internet dependency for library downloads—unblocked a major customer win.
- Earlier: HA cluster management and MySQL clustering/DR on Junos Space.
Senior Software Engineer
Bluecoat Systems
01.2012 - 01.2013
- Cloud Services platform—drove Spring + RabbitMQ POCs and adoption, REST web services, DAO/SQL design, Flex3/ExtJS4 GUI.
Senior Software Engineer
Alert Enterprise
01.2011 - 01.2012
- Alert Controls built from scratch on Spring + Hibernate—DAO layer, scheduled-jobs subsystem, Spring-CXF web services.
Senior Engineer, Product Development
Symphony Services
01.2010 - 01.2011
- Jobvite Source recruiting / ATS.
Software Engineer
Cisco Systems
01.2008 - 01.2009
- Built PSMART (PSIRT vulnerability tracking) and the Feature Tracking System for IOS / IOS XE / IOS XR.
Systems Engineer
Tata Consultancy Services
01.2005 - 01.2008
- SWIFT EAI engine for Clearstream, Fidelity, RJ, FNB.
Education
B.Tech. - Information Technology
SSN College of Engineering, Anna University
Chennai, India04.2005
Skills
- AI / ML & AGENTS
- LangGraph
- A2A protocol
- Claude Agent SDK
- Mem0
- Agent routing
- Automated skill discovery
- Knowledge-graph reasoning
- Config-driven onboarding
- RAG & RETRIEVAL
- Milvus (vector DB)
- Embedding-based retrieval
- Chunking strategies
- Pluggable ingestion/retrieval pipelines
- NVIDIA Triton Inference Server
- INFRASTRUCTURE
- Kubernetes
- Docker
- Nginx
- Spinnaker
- SaltStack
- Terraform
- Helm
- ETCD
- CI/CD
- CLOUD & KNOWLEDGE GRAPH
- AWS (EC2, S3, ELB, MSK, ElastiCache)
- Neo4j
- TigerGraph
- OBSERVABILITY & TRACING
- Prometheus
- Alertmanager
- Grafana
- ELK
- OpenTelemetry (OTEL)
- Jaeger
- Distributed tracing
- Catchpoint (synthetic monitoring)
- SLO/SLI practices
- DATA & MESSAGING
- PostgreSQL
- MySQL
- Cassandra
- Redis
- Oracle
- RabbitMQ (cluster federation)
- JMS/ActiveMQ
- FRAMEWORKS
- Spring
- Hibernate
- REST/CXF
- Jinja
Timeline
Principal Engineer
Apple Inc
12.2017 - Current
Staff Software Engineer
Juniper Networks
02.2013 - 12.2017
Senior Software Engineer
Bluecoat Systems
01.2012 - 01.2013
Senior Software Engineer
Alert Enterprise
01.2011 - 01.2012
Senior Engineer, Product Development
Symphony Services
01.2010 - 01.2011
Software Engineer
Cisco Systems
01.2008 - 01.2009
Systems Engineer
Tata Consultancy Services
01.2005 - 01.2008
B.Tech. - Information Technology
SSN College of Engineering, Anna University