Two architecture disciplines in direct tension.

AI Architect Software Architect

Agent systems up front. Durable software underneath.

AI Architect

Agents. Context. Decisions.

Build the intelligence layer: orchestration, retrieval, tool use, and control.

Agent systems
Context and tool design
Evaluation and guardrails
Knowledge and retrieval architecture
Enterprise AI integration

Collision Zone

Vijay Nirmal

AI Architect | Software Architect | Technical Lead

Tamil Nadu, India

Software Architect

Platforms. Contracts. Scale.

Build the system layer: services, frameworks, resilience, and performance.

Platform foundations
Distributed systems
Performance and resilience
Performance and cost optimization
Cloud modernization strategy

Experience

8+ Years in architecture

Software

30+ Frameworks built

Software

10+ Solutions designed

Software

50x Peak speedup

Leadership

6+ Teams led

Microsoft Agent Framework Agent orchestration

Azure AI Foundry Enterprise AI stack

Software

.NET + Azure Platform architecture

01 / Dual Practice

Architecture across Microsoft platforms and modern web stacks.

From Microsoft Agent Framework and Azure AI Foundry to LangChain, .NET, Node.js, Angular, APIs, and Azure platforms.

Microsoft Agent Framework and LangChain orchestration
Azure AI Foundry and Azure OpenAI workflows
MCP, A2A, and enterprise tool integration
Retrieval, grounding, and knowledge architecture
Evaluation, guardrails, and human approval loops

Signature move

AI systems that blend Microsoft platform depth with practical orchestration and production governance.

.NET, AspNetCore, and Azure solution design
API contracts, microservices, and event-driven integration
Reusable platform foundations across .NET, Node.js, and Azure
Performance, resilience, and cost optimization
Legacy modernization through Web Components and WASM

Signature move

Platforms that combine Microsoft strengths with modern web engineering to scale delivery speed and reliability.

Software

Performance is architecture

Throughput, latency, and cost are design inputs, not cleanup after launch.

Agent orchestration

Design coordinated multi-step flows across Microsoft Agent Framework and LangChain runtimes.

Software

Zero-downtime ETL redesign

Reshape brittle data flows to eliminate recurring downtime and preserve delivery continuity.

Azure AI Foundry

Shape enterprise AI delivery around evaluation, deployment fit, and operational governance.

Software

Event-driven consistency

Use event-driven design with strong consistency where business workflows cannot tolerate drift.

MCP integration

Connect tools, context, and enterprise systems through protocol-driven orchestration patterns.

Software

Proof first, scale second

Prove architecture direction early, then harden the path that survives real usage.

RAG grounding

Use retrieval and knowledge design to ground model responses in current business context.

Software

Platforms compound

Shared frameworks and architecture primitives multiply every team's output.

Human approval loops

Keep critical decisions observable and reviewable with human-in-the-loop checkpoints.

Software

Modernize by evolution

Upgrade real enterprise systems without forcing wasteful rewrites.

Evaluation and guardrails

Measure behavior, constrain risk, and harden AI systems before they become operational dependencies.

02 / Developer Depth

Development depth from framework design to cloud delivery.

Built from the resume's backend, frontend, database, messaging, Azure, DevOps, testing, and container platform experience.

AI engineering and agent delivery

Microsoft Agent Framework, LangChain, and OpenAI
MCP integration and agent workflows
Retrieval, grounding, and evaluation patterns

Backend

.NET, Node.js, and service engineering

AspNetCore, NodeJs, and NestJs
Web API and microservices implementation
Framework and library development

Frontend

Angular and component-based UI delivery

Angular, HTML, CSS, and SCSS
Web Components and UI implementation
Modern experiences inside legacy systems

Data

SQL and NoSQL application data design

SQL Server, MongoDB, and CosmosDB
SQL and NoSQL data modeling
Cross-store read and write model sync

Messaging

Asynchronous and event-driven integration

Kafka, RabbitMQ, Service Bus, and Event Hub
SignalR and real-time communication
Event-driven flows with strong consistency

Infra

Azure, DevOps, and container platforms

Azure tech stack, CloudOps, and Azure CI/CD
Kubernetes, AKS, Istio, and containerization
Azure Pipelines and deployment automation

Architecture

System design and solution shaping

Patterns

Design patterns and reusable abstractions

Delivery

Proof-of-concept to production hardening

Quality

Unit Testing and Test Driven Development

Performance

Profiling, optimization, and scale-aware design

Leadership

Technical guidance and cross-team enablement

03 / Experience

Hands-on leadership across AI delivery, platform architecture, and performance engineering.

The work history is not a list of titles. It is a record of building systems, removing delivery risk, and creating reusable architecture under real product pressure.

Technical Lead

Qualibar

Jul 2025 to present

Designed an AI system and led the team that delivered an agentic application using LangChain and Microsoft Agent Framework.
Reshaped lead ETL architecture to remove a recurring downtime window and support zero-downtime delivery.
Drove the path from architecture design to successful production rollout.

Technical Lead

Neudesic

May 2022 to Jul 2025

Led architecture across 6+ teams while continuing hands-on code contribution.
Designed 10+ reusable libraries and frameworks in .NET and Node.js.
Built high-value proofs including .NET in the browser via WASM, Web Components inside legacy applications, and continuous SQL Server to MongoDB model sync.
Improved critical workloads by up to 50x while remaining cost conscious.

Software Engineer

H&R Block

Jan 2019 to May 2022

Designed and implemented 8+ internal frameworks in .NET.
Built 5+ AspNetCore microservices and Angular user interfaces.
Optimized code paths for over 500% throughput improvement and roughly 60% lower response time.

Backend

AspNetCore, Node.js, NestJS, Web APIs, Microservices

Microsoft Agent Framework, LangChain, MCP, A2A, OpenAI

Frontend

Angular, HTML, CSS, SCSS, Web Components

Data

SQL Server, MongoDB, CosmosDB, Kafka, RabbitMQ, Service Bus, Event Hub

Cloud

Azure, Kubernetes, AKS, Istio, SignalR, Azure CI/CD

04 / Proof

Selected proof points instead of filler portfolio tiles.

These are the kinds of outcomes the work optimizes for: shipping AI systems, compounding platform leverage, and raising the technical ceiling of the teams involved.

Production AI

Agentic system delivery at Qualibar

Led the architecture and team delivery for an agentic application built with LangChain and Microsoft Agent Framework.

Platform Leverage

Frameworks that scaled across teams

Designed reusable libraries and frameworks that removed repetition and accelerated delivery in both .NET and Node.js environments.

Performance

High-impact optimization work

Improved critical application paths by up to 50x in some cases while also reducing cost and operational friction.

Modernization

Bridging old and new architecture safely

Used WASM, Web Components, and hybrid integration patterns to bring modern capabilities into legacy product estates.

Selected Projects

High-signal work beyond client delivery

C# / Recent AIToolkit Focused .NET packages for building AI tooling and integrations. TypeScript / Pinned AzureAISearchExplorer Lightweight GUI for exploring and managing Azure AI Search services. TypeScript / MCP MCPCopilotRelay Generates VS Code extensions from MCP servers by mapping tools and resources to editor capabilities. C# / Performance JsonCraft High-performance .NET JSONPath library built on top of System.Text.Json. TypeScript / MCP WebInteractMCP Turns web applications into MCP servers and expands the practical MCP ecosystem. C# / Official Contribution Microsoft Garnet Contributor Official contributor to Microsoft's high-performance Garnet cache-store project.

Knowledge Sharing

Behind .NET video series

All about Task

Asynchronous Pattern in C#

Basic Understanding

Microsoft Certifications

Cloud and platform credibility

Azure Solutions Architect

Azure Developer Associate

Azure Administrator Associate

Azure AI Fundamentals

Azure Data Fundamentals

Azure Fundamentals

05 / Profile

Profile, community footprint, and contact coordinates.

This portfolio is meant to show the shape of the work. The details below anchor that work in community, publication, and day-to-day engineering practice.

AI Architect Software Architect

Agents. Context. Decisions.

Platforms. Contracts. Scale.

Architecture across Microsoft platforms and modern web stacks.

AI focus