Software Factory Architect · AI Agents · Cloud Platforms
Anji Keesari · Cloud & AI Architect
Published AI researcher · Globee awards judge · San Ramon, CA
Building the AI-powered software factory.
I help enterprises take software from requirement to production, including the applications themselves, with AI agents, a governed engineering platform and secure cloud infrastructure working together. 24+ years of architecture leadership, including regulated financial services.
Read the articlesLet's talkPublished AI researcher · Globee awards judge · San Ramon, CA
24+ yrsarchitecture leadershipFinancial servicesregulated environmentsEnd to endrequirement to production
Free e-books● read online
Free
Building Microservices with Containers
A practical guide to creating microservices with Docker containers, from hands-on experience.
Read online →
FreeIn progress
Building Scalable Kubernetes Infrastructure for Microservices
A practical guide to scalable Kubernetes infrastructure on Azure with Terraform. Chapters are added as they are written.
Read online →The big idea
The software factory, in motion.
Business intent goes in. An orchestrated team of specialist AI agents does the work in parallel, inside guardrails, with people approving at the checkpoints that matter. What happens in production feeds the next cycle.
Intent to impact● live
Business intentgoals · requirements
Orchestrator agentplans · assigns · coordinates
Planstories · options
Buildcode · refactor
Testwrite · run
Securescan · flag
Operatewatch · triage
Governed pipeline
Policy as codeQuality gatesSecurity scansAgent evaluations
Human approves designHuman approves release
Azure + Kubernetesservices · data · identity
Productionobserved & secure
Telemetry & feedback return to intent
Agents work in parallelSpecialists run side by side, coordinated by an orchestrator.
People approve at checkpointsHumans own design, risk and release decisions.
Governed by defaultPolicy, quality and security are built into the path.
What I design for
Goals that matter to leaders
Technology choices are only worth making if they move these four things.
🚀
Faster delivery
A shorter, repeatable path from requirement to production through automation and AI assistance.
🛡️
Trust & governance
Security, compliance and human oversight built into how software and AI agents operate.
📈
Scale & reliability
Applications and platforms that grow with demand and stay available when it matters.
💡
Cost & focus
Less undifferentiated toil, so teams spend their time on what moves the business.
The four layers
Where I go deep
Each layer links to hands-on material for the engineers who build it.
Application Engineering
Requirements turned into working software: designed to be testable and maintainable.
.NET / C#PythonReactNode.jsAPIsMicroservicesSee application work →AI Agents & Agentic Frameworks
Enterprise agents that integrate with existing microservices, safely and observably.
Agent frameworksMCPAzure AI FoundryLLMsExplore agentic AI →Software Factory & Platform
Delivery platforms with automation, policy and security as code.
Azure DevOpsGitOps · ArgoCDTerraformDevSecOpsSee the platform →Cloud Infrastructure
Secure, scalable Azure and Kubernetes foundations built to enterprise standards.
AzureAKSIdentityObservabilityCloud guides →Working together
How I engage
From executive strategy to hands-on proof of concept.
01
Advise
Strategy and roadmap for AI-enabled delivery, translated into language leadership can act on.
02
Architect
Reference architectures and proofs of concept that demonstrate business value before you scale.
03
Enable
Mentoring teams and bridging dev, ops and security so the capability outlasts the project.
Credibility
Practitioner and researcher
Hands-on delivery, backed by peer-reviewed work and independent recognition.
"What excites me most is not just building infrastructure, but reimagining how technology can solve complex business challenges, mentoring teams to reach their full potential."
Read the full profile →2025peer-reviewed papers
Globeeawards judge, reviewing enterprise submissions
SAS · IOASDfellow
IEEEmember
Perspective
Insights & research
Selected thinking, from research to practice.
Research · WIP
Integrating AI Agent Frameworks with Existing Microservices
An adapter-based architecture for enterprises: how to bring agents into the systems you already run.
PlatformAzure DevOps Private Agents in AKS
Self-hosted build agents on Kubernetes.
SecurityPasswordless Access with Workload Identity
AKS pods to PostgreSQL, no secrets.
Building an AI-enabled software factory?
Let's talk about where you are, what's possible and the shortest path to value.
Let's talkConnect on LinkedIn