Writing

Cloud & AI Architecture

Writing by Harish Shenoy on enterprise AI deployment, cloud architecture, engineering leadership, and the practical realities of building systems at scale.

I Gave Every Pull Request Its Own Database
How ephemeral, isolated database environments per pull request eliminate shared-state bugs, speed up CI pipelines, and let teams test schema migrations confidently before they ever reach production.
→
Deploying Lakebase on Azure Databricks with Terraform: A Reusable Framework
A practical guide to provisioning Databricks Lakebase on Azure using a reusable Terraform framework — covering workspace setup, Unity Catalog, storage configuration, and the patterns that make the deployment reproducible across environments.
→
What I Learned Leading the Amazon Aurora Limitless Launch
Leading 80+ people on one of Amazon's top four executive-level strategic priorities in 2023. Lessons in executive alignment, quality automation, and program management at scale from the re:Invent 2023 launch.
→
Five GenAI Architecture Patterns That Delivered Real Enterprise Results
Real patterns from real production deployments: multi-agent crew systems (85% response time reduction), domain-specific RAG (70% accuracy improvement), AI-augmented contact centers, autonomous ops agents, and AI-assisted engineering workflows.
→
How an AI Agent Reduced Network Incident Response Time by 65%
A detailed architecture walkthrough of the autonomous AI agent built for SRE and network operations, integrating Aruba, Cisco, and IPfabric data sources to deliver autonomous root cause analysis and cut MTTR by 65%.
→