Prashanth N.

About Prashanth N.

I build the infrastructure that lets engineering teams ship faster and sleep better - at the scale where a single pipeline decision affects millions of IoT devices. For the past three years at T-Mobile, I\'ve owned DevOps, SRE, and Platform Engineering for the SyncUP IoT platform: a multi-region Azure estate of containers running across AKS and T-Mobile\'s internal multi-cloud Kubernetes platform spanning AWS, Azure, GCP, and on-prem. That means owning CI/CD, cluster operations, DevSecOps gates, observability, incident response, and cost - end to end. The work I\'m most proud of: I built and shipped a multi-agent AI SRE assistant on Claude Code and Azure AI Foundry that cut post-deploy investigation time from 30 minutes to under 90 seconds - a 95% MTTR reduction. Parallel agents collect kubectl output, KQL signals, and Service Bus metrics simultaneously, then route to a reasoning model for root-cause analysis. Engineers without VPN access reach it through a shared MCP server registry. This is the kind of tooling that used to take a team a quarter to build. Beyond the AI work: I migrated the entire SyncUP platform from Azure DevOps to GitLab CI/CD, consolidated the IaC estate, and cut pipeline execution time by 83%- from 12 minutes to under 2. I define SLOs and SLI dashboards for API availability and IoT Hub ingestion latency. I enforce DevSecOps with MR-blocking gates combining SonarQube, GitGuardian, and container CVE scanning. I run Istio mTLS, HashiCorp Vault dynamic secrets, and AD-group-driven RBAC across the full stack. And I\'ve cut cloud spend by 20% while storage accounts dropped 65%. Core stack: Azure (AKS, Event Hub, Service Bus, Key Vault, APIM, App Insights), GitLab CI/CD, Terraform/Bicep, Kubernetes (Istio, KEDA, Strimzi), Prometheus/Grafana/OpenTelemetry, HashiCorp Vault, Python/Bash, and AI tooling (Claude Code, Azure AI Foundry, multi-agent orchestration). CKA · AZ-400 DevOps Engineer Expert · AZ-104 Azure Administrator · MS Computer Science Open to Senior DevOps, SRE, Platform Engineering, and Cloud Infrastructure roles — remote or hybrid.

Experience

1. Senior Cloud Devops Engineer

   T-Mobile

   May 2023 — Present · Seattle, WA, US

   Architected and shipped a multi-agent AI SRE platform on Claude Code and Azure AI Foundry — parallel agents collect kubectl, Application Insights KQL, and Service Bus DLQ signals in real time, then route to a reasoning model for automated root-cause analysis. Post-deploy MTTR dropped from 30 minutes to under 90 seconds (95% reduction), with zero-VPN access via a shared MCP server registry for all platform engineers- Built and own the SyncUP Platform Engineering foundation: migrated from Azure DevOps to GitLab CI/CD, consolidated all service repos under a single template, unified IaC across multiple repositories, and cut pipeline execution from 12 minutes to under 2 minutes (83% reduction) — increasing deployment frequency by 60% and reducing lead time by 40%- Defined SLOs, SLI dashboards, and error budget tracking for API availability and IoT Hub ingestion latency using Prometheus, Grafana, and OpenTelemetry OTLP — enabling proactive toil reduction and data-driven reliability decisions across all platform services- Enforced zero-tolerance DevSecOps by wiring SonarQube, GitGuardian, and container CVE scanning into an MR-blocking gate — clearing 40+ PacBot, GitGuardian, and CVE findings across Linux AKS workloads, enforcing AppArmor profiles, rotating all PKI certs via Keyfactor/Azure Key Vault, and building HashiCorp Vault dynamic secrets with no long-lived credentials anywhere in the pipeline- Operated Kubernetes fleet across AKS and Conducktor/TKE (T-Mobile\'s multi-cloud platform spanning AWS, Azure, GCP, on-prem) — owning major-version upgrades for KEDA, kube-prometheus-stack, Strimzi Kafka Operator, and OpenTelemetry Collector across dev/stg/ppd/prd with 48-hour burn-in periods- Cut cloud spend by 20% and storage accounts by 65% through autoscaling, reserved instances, Azure Cost Management tagging, and systematic decommissioning — while profiling critical services to cut response times by 30%.

Education

• University of Central Missouri

  Masters, Computer Science