How to standardize LLM tool calling, schema discovery, and enterprise context integration using the Model Context Protocol in C# and .NET 8.
I am Bhushan Kadam
With 13+ years of experience building scalable enterprise distributed systems using .NET & Microsoft Azure, I work at the intersection of production architectures, Model Context Protocol (MCP), and agentic engineering workflows.

Background • Philosophy
About Me
I have spent the last 13+ years designing, building, and operating enterprise software. Over my career at Dell Technologies, JustAnswer, Allscripts, Wipro PARI Robotics, and Ignite Solutions, I have worked across the full lifecycle of complex systems—from real-time robotics guidance and healthcare software modernization to cloud-native microservices on Microsoft Azure.
In recent years, I have focused on applied AI and agentic software engineering. I am passionate about how LLMs and agent frameworks elevate developer productivity and application capability. However, I remain firmly grounded in production engineering reality: systems must be deterministic, secure, maintainable, and rigorously verified.
Alongside engineering, I regularly publish architecture guides and .NET deep dives on my blog, Engineering Blog .
My Experience • Career Timeline
13+ Years of Building Production Systems
A chronological track record of my work across enterprise distributed backends, cloud-native modernization, and applied AI systems.
Principal Software Engineer
Current RoleI architect and deliver mission-critical enterprise platforms and distributed systems on Microsoft Azure using .NET and cloud-native patterns, while driving modern agentic coding and spec-driven engineering practices across engineering teams.
- —Architected scalable distributed backend systems utilizing .NET and Microsoft Azure event-driven microservices.
- —Established modern spec-driven development practices integrating AI agents for accelerated delivery and automated verification.
- —Engineered resilient messaging patterns with Azure Service Bus, guaranteeing at-least-once processing and robust fault recovery.
- —Championed observability, automated integration testing, and CI/CD best practices ensuring high platform availability.
Lead Software Engineer
I led engineering efforts for high-traffic customer-facing web platforms and automation infrastructure, combining .NET backend services, React frontends, Python automation, and early production LLM integrations.
- —Spearheaded integration of LLMs into core workflows utilizing function/tool calling and Retrieval-Augmented Generation (RAG).
- —Modernized monolithic services into containerized microservices deployed on Docker and Kubernetes.
- —Mentored senior and mid-level engineers, instituting code review standards, TDD, and automated test coverage.
- —Optimized high-throughput API endpoints to decrease latency and increase system throughput under peak loads.
Senior Software Engineer
I engineered and modernized mission-critical healthcare software applications, ensuring high reliability, patient data privacy, and seamless interoperability across clinical workflows.
- —Modernized legacy healthcare enterprise applications to modular .NET microservices and Microsoft Azure cloud services.
- —Optimized SQL Server database queries and transactional bottlenecks for high-concurrency electronic health record systems.
- —Built rich desktop interfaces using WPF with MVVM architecture, ensuring smooth data binding and responsiveness.
- —Automated comprehensive regression and unit test suites, significantly lowering post-release production defects.
Software Engineer
I developed real-time software systems for industrial automation, high-density automated car parking, and Automated Guided Vehicles (AGVs).
- —Authored sub-second real-time control algorithms for multi-vehicle AGV pathing and collision prevention.
- —Engineered hardware-software communication layers bridging PLC sensors and centralized supervisory software.
- —Delivered turn-key automated storage and parking systems for commercial and manufacturing facilities.
Software Engineer
I engineered client web applications, database layers, and backend services utilizing C#, ASP.NET, and modern web technologies during my early engineering career.
- —Designed and implemented web features, database schemas, and RESTful API endpoints for client applications.
- —Collaborated with engineering teams to ensure clean code practices, bug resolution, and high reliability.
- —Implemented responsive UI components and data binding with backend SQL databases.
Featured Work • Open Source
Architectures, Tooling & Systems I've Built
ASP.NET Core Playground
Production-ready architectural patterns & best practice samples for modern .NET 8+
A curated collection of modern, production-grade ASP.NET Core samples demonstrating clean enterprise architecture patterns for common real-world development scenarios.
Eliminates guesswork in enterprise .NET apps by providing concrete, battle-tested reference implementations for strongly-typed configuration, secure authentication, and decoupled abstractions.
Model Context Protocol (MCP) Enterprise Suite
Connecting AI coding agents to enterprise APIs, databases & telemetry via standardized MCP tools
A Model Context Protocol server and client implementation designed for enterprise software engineering. Enables AI agents to safely query database schemas, inspect API contracts, and run validated operational checks within strict permission boundaries.
Replaces fragmented, bespoke LLM tool plugins with standard Model Context Protocol servers, preventing prompt leakage and unauthorized data access while giving agents rich codebase context.
Spec-Driven Agentic Engineering Workflow
Orchestrating requirement specifications to multi-step agent coding & automated verification
An automated framework that bridges high-level requirements into formal technical specifications, executes code generation across multi-file repositories using AI agents, and enforces automated regression gates before code review.
Eliminates hallucination and drift in AI-generated code by anchoring agents to explicit interface contracts and automated Playwright / NUnit test suites.
Enterprise RAG & Function Calling Orchestrator
Production RAG engine with hybrid semantic retrieval and transactional tool dispatching
A retrieval-augmented generation engine built for complex enterprise knowledge domains. Combines vector embeddings with BM25 keyword search, cross-encoder re-ranking, and deterministic tool calling into enterprise backend systems.
Ensures factual accuracy and citation transparency in LLM answers while providing an auditable execution layer for transactional tasks.
My Engineering Perspective
How I Apply AI & Agentic Coding in Practice
The true power of AI in software development isn't just generating code snippets—it's executing coherent, multi-step engineering workflows with automated verification.
Spec-Driven Development
Requirements and architectural contracts (OpenAPI schemas, database entities, acceptance criteria) are established before coding begins. LLMs and agents operate within unambiguous boundaries, preventing drift and hallucination.
Model Context Protocol (MCP)
Rather than pasting ad-hoc snippets into chat interfaces, AI coding agents interact through standardized MCP servers. This enables agents to query live schemas, inspect repository trees, and execute validated diagnostics within strict security boundaries.
Agentic Multi-File Workflows
Moving far beyond simple autocomplete, coding agents execute coherent multi-step tasks across domain layers—orchestrating command handlers, DTOs, migrations, and test suites in concert.
Automated Verification Gates
AI output is never assumed to be correct without proof. Automated headless test suites (NUnit, Playwright), static analyzers, and security scans provide deterministic proof that generated implementations fulfill the specification.
My Writing • Architecture Notes
Technical Blog & Publications
Moving beyond IConfiguration strings: implementing strongly-typed options, startup validation, and decoupled configuration in enterprise microservices.
Why senior engineering judgment and formal interface specifications are essential to harness AI coding agents without code drift.
Eliminating dual-write failures between relational databases and Azure Service Bus using the Transactional Outbox pattern and background publishers.
A comprehensive architectural evaluation of when to adopt microservices versus retaining a modular monolith in enterprise platforms.
A step-by-step guide to using the Hangfire library to schedule, execute, and monitor background tasks in ASP.NET Core applications.
How to implement production-grade JSON Web Token (JWT) authentication, claims-based access control, and authorization in ASP.NET Core.
Practical strategies for leveraging IMemoryCache in ASP.NET Core to dramatically reduce database bottlenecks and accelerate response times.
An engineering deep dive into asynchronous programming in C#, Task state machines, avoiding deadlocks, and preventing thread exhaustion.
Master the fundamentals of Language Integrated Query (LINQ) in C#, exploring query syntax, lambda expressions, and deferred execution.
Deep dive into advanced LINQ operations in C#: complex joins, GroupJoin, Zip, Aggregate, and memory allocation pitfalls in production.
A comprehensive exploration of C# delegates and events, decoupling system components using clean publisher-subscriber communication.
Step-by-step tutorial on architecting an end-to-end AI chatbot combining a React frontend, Python API backend, and streaming LLM inference.
How to package, optimize, and deploy ASP.NET Core applications into minimal, secure Docker containers for production environments.
Evaluating modern state management paradigms in React: when to use local state, Context API, Redux Toolkit, or lightweight Zustand stores.
Practical techniques for reading and writing large Excel spreadsheets in C# without Microsoft Office dependencies using managed libraries.
A robust approach to sending transactional HTML emails in ASP.NET Core using MailKit, secure credentials, and background queueing.
Optimizing JSON processing in modern .NET: utilizing System.Text.Json, custom converters, source generators, and zero-allocation techniques.
Technical Competencies
My Core Technology Stack
Backend & Cloud Architecture
- C# / .NET 8 & 9
- ASP.NET Core
- Microsoft Azure
- Event-Driven Microservices
- Azure Service Bus
- CQRS & MediatR
- REST APIs & WebSockets
- Clean Architecture / DDD
AI & Agentic Engineering
- Model Context Protocol (MCP)
- Spec-Driven Development
- Agentic Coding Workflows
- LLM Tool & Function Calling
- Retrieval-Augmented Generation (RAG)
- Semantic Kernel & AI SDKs
- AI Developer Tooling
Data & Storage
- SQL Server / Azure SQL
- Entity Framework Core
- PostgreSQL & pgvector
- Redis Caching
- Azure Blob Storage
DevOps & Quality Assurance
- Docker & Kubernetes
- GitHub Actions & CI/CD
- Azure DevOps
- NUnit & Automated Testing
- Playwright E2E Testing
- App Insights & Observability
Get in Touch
Interested in discussing distributed systems, .NET architecture, or AI-native development?
I am always open to high-impact technical conversations, architecture advisory, or engineering collaboration.