WHAT I'M UP TO.
A public log of what I'm actively building, learning, and the technical questions currently keeping my mind occupied.
Active Systems & Platforms
AI-Powered Workflow Systems
Building agentic execution systems that utilize Model Context Protocol (MCP) servers and headless Playwright automation to execute reliable multi-step cloud tasks.
ShipOps Knowledge Platform
Publishing in-depth production guides on Azure landing zones, Terraform state isolation, and DevOps automation best practices.
Personal Engineering Command Center
Crafting tanmaytrivedi.me using the 'Engineering Noir' design system, Next.js static generation, and sub-100ms edge performance.
Architectural Research Areas
Advanced Azure Architecture
Zero-trust multi-region landing zones, advanced ExpressRoute traffic engineering, and private network perimeter design.
Cloud-Native AI Infrastructure
Designing cost-efficient compute runtimes for agentic workloads using Azure Container Apps scale-to-zero and GPU-enabled nodes.
Kubernetes & Service Mesh
Hardening AKS clusters with Cilium eBPF networking, Istio mutual TLS, and automated Argo CD GitOps pipelines.
Questions I'm Exploring
Open engineering questions with non-trivial trade-offs that I am actively testing in production and staging environments.
How much infrastructure does an AI system actually need?
Balancing heavy microservice orchestrations versus lean serverless container runtimes with scale-to-zero compute.
When should a workload use Container Apps vs App Service?
Evaluating cold-start latency, KEDA event-driven scaling, and networking boundaries for production microservices.
How do you design reliable, deterministic AI workflows?
Preventing hallucinated parameter executions through rigorous schema validation (Zod/Pydantic) and MCP boundaries.