Available for architecture engagements

Gowrishankar, AI Solution Architect

I design AI systems that survive contact with production. I design AI, cloud and data platforms for enterprises — then write about what actually worked.

View case studiesAbout my approach
EMR-scale microservice migrationsAI-native dev pipelinesFull-stack across web, cloud & AI

Explore

A living map of everything I've written

Every node is a topic pulled straight from the blog. Click one to see the posts, or search — the graph reacts as you type.

AIArchitectureDecisionMakingC4 ModelOrg DesignAI GovernanceDistributedSystemsCareerTechnicalDebtTeamTopologiesStrangler FigMigrationStrategyStacey MatrixPostel's LawTool CallingAgentSecurityWeb AgentsOODA LoopMCPA2AInverse ConwayManeuverAgentProtocolsFallaciesEvaluationTestingCQRSEventSourcingContextEngineeringAgent MemoryCAP TheoremDocumentationDomain-DrivenDesignSystemBoundariesWardleyMappingStrategyComplianceCynefinStanfordAI IndexIndustryTrendsRAGLLMLLMOpsMLOpsSolutionArchitectureArchitecturePlatform

AI & Agentic Engineering

  • LLM Dev Pipelines (Cursor)
  • Claude Code Agentic Skills
  • AI Data Enrichment
  • RAG-Assisted Workflows

Backend & Architecture

  • Next.js & NestJS
  • Node.js & Express
  • Microservice Migration
  • Strangler Fig Pattern

Cloud & Messaging

  • AWS (EC2, Lambda, DynamoDB)
  • RabbitMQ
  • BullMQ & Redis
  • Docker & CI/CD

Frontend & Mobile

  • React.js & React Native
  • Redux Toolkit / RTK Query
  • Tailwind & Material-UI
  • Expo / Android / Xcode

Data, Testing & Security

  • MongoDB & MySQL
  • Jest & RTL
  • Grafana BI
  • Security & Vuln. Assessment

Selected work

Architecture that shipped, not just diagrams that shipped

A sample of platforms I've designed end-to-end — full case studies in the portfolio.

mConsent 2.0

AI Solution Architect

2025

Architected and led the phased, strangler-fig migration of a legacy PHP EMR platform onto a Next.js/NestJS microservice stack — designing the infrastructure and technical architecture from scratch, and leading a small team through it. RabbitMQ handles inter-service communication; BullMQ and Redis drive the cron-based activity-logging pipeline, backed by MongoDB. Alongside the migration, I led the team's shift to an AI-native development pipeline — Cursor and a custom Claude Code skill system (requirement-interview → clean → verify → commit end to end).

  • A project scoped for ~2 years shipped in significantly less time
  • Strangler-fig phased cutover with zero migration downtime
  • AI-native dev pipeline adopted end-to-end across the team
Next.jsNestJSRabbitMQBullMQRedisMongoDBCursorClaude Code

LeadOS

AI Solution Architect

2025

Designed a lead-generation platform that combines automated NPI registry data enrichment with an AI layer, feeding directly into a sales pipeline built for the lead-gen and SDR teams — from raw registry data to a qualified, sales-ready lead with no manual handoff.

  • Automated NPI registry enrichment that previously required manual lookup
  • Sales pipeline integrated by default for automated SDR workflows
  • Built and owned by the lead-gen team without engineering hand-holding
Next.jsNestJSMongoDBAI Data Enrichment
View all case studies →

Latest writing

Notes from designing AI systems

Architecture patterns, trade-offs, and lessons from production — not tutorials.

·8 min read

AI Solutions Architect vs AI Engineer: C4 Already Solved This Confusion

A 2026 career guide can't cleanly separate Data Scientist, ML Engineer, AI Engineer and AI Solutions Architect. The C4 model already named these four levels — just not by job title.

CareerC4 ModelAI Architecture
·7 min read

The Technical Debt Quadrant: The Only Debt That Should Scare You Is the Debt Nobody Chose

Martin Fowler's reckless/prudent × deliberate/inadvertent grid, and why the quadrant an AI system's debt falls into matters more than the debt itself.

Technical DebtDecision MakingAI Architecture
·7 min read

Team Topologies: The Org Chart Is an Architecture Decision

Four team types, three interaction modes, and why an AI platform's biggest bottleneck is usually the team boundary, not the code.

Team TopologiesOrg DesignAI Architecture
Read all posts →

Designing an AI or platform architecture?

I take on a small number of advisory and architecture engagements each quarter.

Get in touch