Building AI infrastructure and cloud platforms at scale. 13+ years across embedded systems, networking, cloud-native platforms, and AI/ML infrastructure. MS Computer Science, USC.
I'm a staff software engineer with 13+ years of experience across the full stack of modern infrastructure — from embedded firmware and network drivers to cloud-native platforms and AI agentic systems.
My foundation is in systems programming. Early in my career I built hardware drivers, network protocol stacks, and platform software for telecom equipment — work that demands precision and an intimate understanding of how software and hardware interact.
Over time I moved up the stack: cloud application development at a major networking company, multi-cloud SaaS platform engineering, and most recently leading AI infrastructure — building agentic systems that integrate enterprise knowledge sources and automate complex operational workflows.
I care about building things that are reliable, observable, and genuinely useful. I write about engineering and ideas on my Views blog.
Staff Software Engineer, Santa Clara, CA
University of Southern California, 2014–2016
BMS College of Engineering, Bangalore, 2008–2012
AI agentic infrastructure, SaaS platform architecture, LLM integrations
13+ years across AI, cloud platforms, networking, and systems.
Redesigning a large-scale SaaS platform architecture for improved scalability, performance, and resource efficiency. Leading AI agentic infrastructure development — building systems that provide automated assistance and debugging by integrating knowledge from multiple enterprise data sources. Establishing a production engineering practice focused on AI-driven customer issue resolution.
Led global backend AI infrastructure and developer platform engineering. Scaled foundation model integrations on Kubernetes, built agentic AI systems, and reduced search latency significantly. Built data pipelines using OpenTelemetry and Snowflake to generate training data for ML models. Architected multi-cloud, multi-tenant SaaS platform using Kubernetes controllers and Terraform. Grew and managed a cross-geo Developer Experience team; redesigned CI/CD and internal tooling used across the entire engineering organization.
Architected and led development of cloud-native management applications on AWS for router and datacenter infrastructure. Built backend frameworks exposing Python APIs for controllers, UIs, and test automation. Designed model-driven telemetry pipelines for real-time traffic and configuration data. Built traffic flow extraction and visualization tooling using Kafka, Elasticsearch, and Kibana. Led the team's CI/CD build system and container-based deployment workflows. Also contributed EVPN bug fixes and MPLS-over-GRE programming on Broadcom-based ASICs.
Developed SDK APIs for a datacenter network switching fabric, part of a software-defined networking architecture. Focused on API design and integration across the control and data plane.
Built telemetry framework APIs to capture analytics and troubleshooting data from network virtualization deployments. Data was used as a feedback signal to improve feature usability in production customer environments.
Developed platform APIs and cloud platform software on OpenStack and VMware. Built storage connectivity features for geo-redundancy and disaster recovery. Wrote hardware drivers for DHCP, FPGA access, Ethernet switching (Broadcom), SerDes, and CPRI interfaces for embedded telecom platforms running on Linux and VxWorks/RTOS.
Languages, platforms, and tools across 13+ years of engineering.
Highlights from across my career.
Designed and built agentic AI systems that integrate knowledge from multiple enterprise data sources to provide automated assistance and debugging. Infrastructure handles orchestration, retrieval, and multi-step reasoning at scale.
Co-architected a multi-cloud, multi-tenant SaaS platform using Kubernetes controllers and Terraform. Scaled internal datacenter 30x, managing infrastructure for 100k+ monthly workloads.
Built a real-time traffic flow extraction and analysis system for network devices. Data flows from device netflow through Kafka topics into Elasticsearch, visualized in Kibana with custom dashboards.
Built and led a cross-geo Developer Experience team delivering internal tooling, self-service infrastructure, CI/CD redesign with Bazel and ArgoCD, and artifact security scanning and signing across a large engineering org.
Wrote low-level hardware drivers for DHCP, FPGA access, Ethernet switching, SerDes, and CPRI interfaces on embedded telecom platforms. Managed Linux kernel migrations and BusyBox customization for production radio equipment.
This site's Views section is powered by Claude AI. Raw thoughts go in, structured posts come out. A GitHub Actions workflow calls the Claude API and publishes directly to this site — zero manual editing needed.
Thoughts on AI, infrastructure, engineering culture, and building things at scale — shaped with Claude. I prompt, Claude writes, it publishes here automatically.