About Ajin
I'm Ajin Baby — an AI Platform & Cloud Infrastructure Architect and a Staff Software Engineer at a global aviation-software company, where I own the architecture for how AI agents and LLM applications connect to internal systems and data, including the company-wide MCP gateway they go through. I've been on this platform for nearly a decade — most of it as a Computing Architect — so I've watched, and built, the shift from runbooks and dashboards to agents, MCP surfaces, and context graphs up close. In an industry where reliability and safety aren't optional, the job has always come down to the same thing: cloud, data, and AI foundations teams can ship on safely.
Over 15+ years I've settled into the space where AI platforms, agentic infrastructure, and cloud platform engineering meet. In production that has meant Azure OpenAI and Azure AI Foundry, agentic workflows with vector databases and orchestration frameworks, Model Context Protocol (MCP) services, and Neo4j graph databases driving GraphRAG and context-aware retrieval. It has also meant the unglamorous, load-bearing work: ETL pipelines and Databricks Asset Bundles; AWS and Azure across compute, storage, networking, serverless, and directory services; Kubernetes, Helm, and service mesh; CI/CD on Terraform, Ansible, Azure DevOps, GitLab, and GitHub; and identity across Okta, Auth0, Entra ID, PingFederate, and PingOne. What I care about most is the reliability, governance, and security patterns AI systems need when they run in regulated, safety-critical environments.
I didn't start in architecture — I started by building companies. As a 2x founder, I co-founded Veeble SoftTech and helped grow it from 3 people to 20, and from $500 in capital to ~$500K ARR across 110+ countries, while consulting 50+ companies through their cloud migrations. Alongside it, I co-founded and ran ExTravelMoney as CEO, a reverse-auction marketplace for India's retail foreign-exchange market that won the 2016 Dr. A.P.J. Abdul Kalam Startup Award. That founder's instinct — start simple, ship, then scale only when you have to — still shapes how I build.
I hold an MS in Management Information Systems from the University of Arizona and am pursuing an MS in Artificial Intelligence at Indiana Wesleyan University. I've earned Microsoft's Azure AI Engineer Associate and Azure Administrator Associate, AWS Solutions Architect – Associate, Neo4j Certified Professional, and CNSS credentials. I'm also a 15+ year IEEE volunteer and a recipient of the 2013 IEEE MGA GOLD Achievement Award (Senior Member application in progress). This site is where I write and build in public for engineers working on AI-enabled cloud platforms, MCP and agentic systems, and reliability patterns for the next generation of infrastructure. The domain says cloudandsre because reliability is the discipline that holds all of this together. I live in Parker, Colorado.
The Book
Self-Healing Infrastructure: Building Autonomous Cloud Systems with AI is my practitioner's guide to AI-enabled cloud platforms in regulated industries: MCP in production, closed-loop remediation, and bounded autonomy, with seven working labs. It's free, 262 pages, and archived with a citable DOI.
Research
I publish preprints on reliability engineering for AI-enabled infrastructure, each with runnable companion code. They're self-published and not peer reviewed.
- From Incident to Failing Build: Incident-Derived Regression Testing for AI-Agent Guardrails
- Stateless Protocol, Stateful Problem: Durability, Retry, and Retention Policy for MCP Tasks
- Error Budgets for Autonomy: SLO-Driven Authority Management for Autonomous AI Operators
- Stable by Design: A Control-Theoretic Account of AI-Driven Self-Healing Remediation Loops
Open Source
Tools and reference implementations for running agents reliably: chaos testing for MCP tool calls (mcp-chaos), a replayable corpus of real agent failures (agent-incident-corpus), runtime spend limits (agent-spend-guard), context graphs built from a Backstage catalog (catalog-context-graph), and everyday SRE tooling like incident-scribe and alert-explainer. The rest are at github.com/ajinb.
Writing
167 posts so far: fundamentals primers, breakdowns of real incidents, and field notes from building agentic systems.
Connect
For talks, paper reviews, or research collaboration, LinkedIn is the best way to reach me.