Scott Murphy — Portfolio
Senior .NET / Azure Engineer & Architect · Production Agentic AI Systems · Independent AI Engineering (Freelance)
Architecting enterprise .NET/Azure systems — now building governed, production multi-agent AI. I ship real systems, not slides.
Professional Summary
Senior .NET/Azure engineer specializing in enterprise modernization and AI engineering. Deep experience with cloud-native architectures, Clean Architecture, and microservices on Microsoft Azure, plus production agentic-AI systems — multi-agent orchestration, RAG, and MCP tool integration built on the Azure AI stack (Semantic Kernel, Azure AI Search, Azure OpenAI).
Since 2025 I've built and shipped production agentic-AI systems on independent freelance / contract engagements — multi-agent orchestration, RAG pipelines, MCP tool integration, and the Microsoft Azure AI stack with Semantic Kernel across .NET, Next.js, and React Native, deployed to live environments. The execution depth of a senior engineer with hands-on command of modern AI engineering, applied to enterprise-grade problems.
Experience
MiChoice Technology — Software Engineer (Freelance / Contract)
Jan 2024 – Present
Tasked with creating an agentic bidding system (MiBids) to assist in the bid process.
.NET 10 / ASP.NET Core / EF Core (SQL Server); confidence-gated answer matching with a human approval gate; model-agnostic LLM layer (Azure OpenAI, OpenAI, or local); React and Next.js front ends.
Carborundum.ai — Independent AI Engineer (Freelance / Contract)
Jan 2025 – Oct 2026
Tasked with creating an agentic field application (Carbon) to supplement their existing office system (Silicon).
React Native / TypeScript; token-authenticated REST to Silicon (Next.js, Prisma/PostgreSQL); offline queue; job completion updates ERP inventory and low-stock alerts; voice-assistant prototype.
Senior Technical Consultant, App Modernization Consulting (US) · Perficient
Nov 2025 – Apr 2026
Cloud-native .NET 10 microservices & enterprise API engineering for Lockton (insurance), supporting binder management for 100+ accounts. Binder Management API on Clean Architecture; QuoteCompare AI for document extraction & premium pipelines; M-Files integration; Azure Redis distributed caching; Salesforce (Forza.Net/SOQL) microservices; CI/CD with NUnit/Moq.
Senior Software Engineer · Loomis US
Oct 2022 – Feb 2025
Led modernization of Loomis Direct from legacy .NET to .NET Core + React SPA. Drove Azure cloud adoption and optimized CI/CD, cutting deployment time by 30%.
Software Engineer (Contract) · TransCore
Oct 2019 – Aug 2022
Modernization solutions for state-government systems using Angular, C#, and Azure-hosted SQL. Converted multi-threaded legacy data processing into scalable, maintainable modules.
Senior Software Engineer · Roquemore & Roquemore
Jun 2015 – Oct 2019
Modernized internal WPF and MVC applications. Led third-party integrations including Equifax, TransUnion, and CarFax.
Software Developer · Automated Integration
Feb 2011 – Apr 2015
Revitalized a legacy ERP/dispatching system with modernization-ready enhancements. Built field-service mobile apps in C# and SQL Server.
Skills & Stack
Production Agentic AI Systems
Multi-agent orchestration, MCP tool integration, LLM prompt architecture, RAG pipelines, embedded AI advisory layers (miAgentic), governance guardrails, human-approval gates — built and deployed (MiChoice, Interview Ready) and in public demo repos. Claude API; Microsoft Azure AI stack (Azure OpenAI, Azure AI Search for RAG, Semantic Kernel) as the enterprise bridge.
Core .NET Backend & Modernization
C# / .NET 10 / ASP.NET Core 8, Clean Architecture, DDD, EF Core, SignalR streaming, REST APIs, microservices, modular-monolith refactoring
React 19 / Next.js, TypeScript, Tailwind, SPA migration
React 19 / Next.js, TypeScript, Tailwind, SPA migration
Enterprise Delivery & Cloud
Azure App Services, Azure Redis, Azure DevOps CI/CD, Azure Storage
Salesforce (Forza.Net/SOQL), M-Files, Stripe, integration patterns
Agile/Scrum, stakeholder alignment, testing (NUnit, xUnit, Moq)
Salesforce (Forza.Net/SOQL), M-Files, Stripe, integration patterns
Agile/Scrum, stakeholder alignment, testing (NUnit, xUnit, Moq)
Education & Certifications
University of Houston, College of Technology
Professional Scrum with Kanban (PSK I)
Lean Six Sigma Yellow Belt
Professional Scrum with Kanban (PSK I)
Lean Six Sigma Yellow Belt
Proof of Agentic Work
Agentic AI systems I’ve designed, built, and run — not slideware.
Each one below is real: deployed and running, not mocked up.
MiChoiceSITE
Tasked with creating an agentic bidding system (MiBids) to assist in the bid process.
.NET 10 / ASP.NET Core / EF Core (SQL Server); confidence-gated answer matching with a human approval gate; model-agnostic LLM layer (Azure OpenAI, OpenAI, or local); React and Next.js front ends.
getmichoice.com
Carborundum.aiSITE
Tasked with creating an agentic field application (Carbon) to supplement their existing office system (Silicon).
React Native / TypeScript; token-authenticated REST to Silicon (Next.js, Prisma/PostgreSQL); offline queue; job completion updates ERP inventory and low-stock alerts; voice-assistant prototype.
carborundum.ai
Multi-agent orchestrationVIDEO
A lead agent decomposing work and dispatching it to 8 specialized agents under a human-approval gate — recorded running live, not a mockup.
youtu.be/zmoMVRP447I
Ground Truth — site-intelligence briefs
A per-parcel property-intelligence brief generated from live public data — real costs, taxes, HOA bylaws, zip-level crime, schools, hospitals, aerial imagery, and a cinematic flythrough. Designed, built, and shipped end-to-end.
Interview Ready
Give it a company and a job description; it researches the role and builds a personalized interview study guide — a research-to-guide pipeline, deployed and running end-to-end.
Writing
Agentic Development Is Still Sunday SauceNEW
The patterns behind agentic AI aren't new — a SKILL.md is a class, a prompt chain is a foreach, MCP registration is dependency injection. C# fundamentals, renamed.
linkedin.com/posts/scottroymurphy_agenticai…csharp
Agentic engineers still need to be full-stack. Not in their hands. In their heads.
Why agentic AI raises the bar for architectural and systems thinking — the fundamentals still live in the engineer's head, even when the model writes the code.
linkedin.com/posts/scottroymurphy · 7473052740445687808
The Legacy Migration Playbook
A field-tested playbook for modernizing legacy enterprise systems with AI — assessment, decomposition, and delivery.
linkedin.com/posts/scottroymurphy_the-legacy-migration…
The gap between "our app works" and "our app works with AI" is smaller than you think
A single-session case study: adding AI-powered nutrition lookups to a 12-year-old .NET WebForms app (MiMealOrders) via Assay → Blueprint → Transmute → Deliver.
linkedin.com/posts/scottroymurphy_ai-dotnet-legacymodernization…