Divyansh
Srivastav
DevOps Architect · Kubernetes Platforms × Agentic AI
Architect @ Qualitest · 2026 – present · 11 years across Azure, AWS, GCP
I design and architect cloud infrastructure and Kubernetes platforms across Azure, AWS, and GCP — focused on systems that are scalable, secure, highly available, and cost-efficient by design. I rely on GitOps, Infrastructure as Code, and strong observability practices to ensure these platforms can evolve without accumulating operational debt.
Over the past eleven years, I've worked across three cloud ecosystems, delivered infrastructure for regulated industries including healthcare, and owned architecture decisions across the full lifecycle — from design to production — while leading cross-functional DevOps teams of up to 22 engineers. In one engagement, this meant taking a multi-tenant SaaS platform from zero to production-ready in 45 days.
My current focus is applying agentic AI to reduce the operational toil that still requires human intervention — automated deployment validation, infrastructure drift detection, and intelligent incident triage. Building toward platforms that handle the routine work themselves so engineers can spend their time on the problems that actually need judgement.
11yrs
industry experience
3clouds
Azure · AWS · GCP
CKA
Certified K8s Admin
TFAssoc
Terraform Assoc
AI
Agentic AI · DL.AI 2026
What I do
Capabilities
Platform Engineering
Kubernetes platforms built for autonomous operation — GitOps-driven, Terraform-managed, with RBAC and multi-tenancy designed in from day one.
Cloud Architecture
Multi-region, multi-cloud infrastructure designed for real failover — not just theoretical resilience. Cost-optimized and auditable by default.
AI-Augmented DevOps
LLMs wired into CI/CD pipelines for Terraform plan review, policy validation, and incident triage — practical AI that reduces toil, not demos.
Latest Writing
Featured Articles
Building a Multi-Agent LLM Debate Engine with Python
Architecture decisions, 4 LLM provider integration, and lessons learned building Synapse — a 45-file, 12,000-line multi-agent debate system with a 5-phase pipeline and 306 tests.
AI-Augmented DevOps: Using LLMs to Review Terraform Plans and Predict Deployment Failures
How AI agents are transforming infrastructure operations—from automated Terraform plan reviews to predicting deployment failures before they happen, cutting incident response time by 60%.
Migrating off AAD Pod Identity on Multi-Region AKS
AAD Pod Identity hit end-of-support in late 2024. The migration to Workload Identity on a multi-region AKS estate is straightforward at the YAML level — and brutal in the details. Two failure modes from cutover, the architecture that works, and what to budget for.