AI Technologies

AI Strategy and Engineering Enablement

In my employment work, I led architecture work on AI-assisted engineering, AI-generated testing, agent orchestration, and governance. I helped guide Software Factory adoption and served as the lead for engineer-focused AI enablement and technical implementation within the Education AI strategy. My contributions included AI artifact review models, reuse and skills governance, and contributions to Foundry and Bedrock evaluations.

Explore my employment AI accomplishments. The private product work below is a separate part of my technical experience.

Reusable AI Authoring Foundations

In the QA framework I developed, the delivered AI-assisted authoring layer includes schema-validated YAML workflow specifications, driver-neutral application maps, Playwright MCP application discovery, and specification-to-test traceability through NUnit workflow identifiers.

AI and MCP participation ends at authoring. Tests execute through the existing .NET, API, Selenium, and Playwright abstractions without an AI-service dependency. Automated test generation, coverage analysis, repair, and impact analysis remain planned extensions of this framework.

The developer toolbox also provides curated engineering instructions, roles, prompts, and skills through a canonical shared library, with synchronized GitHub Copilot, Codex, and Claude Code adapters. These reusable assets support engineering teams alongside the architecture, governance, and product work described here.

Explore the delivered QA authoring capabilities and the developer toolbox. See the engineering enablement practices.

AI Product and Platform Work

AI is now part of both my product thinking and my engineering workflow. In current private product work I am building AI-enabled experiences around conversational coaching, prompt-driven workflows, content generation, and automation. That work includes modern .NET web applications, chat-oriented user experiences, provider abstraction, observability, and careful system boundaries around data, validation, and supportability.

LLM, Voice, and Media Orchestration

My recent private project work goes beyond simple prompt experiments. It includes multi-provider LLM integration, AI chat experiences, text-to-speech orchestration, image and video generation pipelines, continuity and quality gates, and configurable provider patterns designed so the underlying AI services can evolve without forcing a full application rewrite.

Modern AI Stack Experience

The technologies I am currently working with in this space include .NET 10, Blazor, ASP.NET Core, SignalR-style real-time UX, Azure OpenAI and Azure AI services, Azure Speech, Redis, Cosmos, SQL Server, OpenTelemetry, Seq, Application Insights, and feature-flag driven rollout patterns. My public seq-mcp project complements that by showing how I think about AI tooling, structured integrations, and operational diagnostics for agent workflows.

AI-Assisted Engineering Workflow

I actively use GitHub Copilot, Codex, and OpenAI ChatGPT as part of my day-to-day workflow. I use them differently depending on the need: inline coding assistance and scaffolding with Copilot, repository-aware task execution and refactoring help with Codex, and broader problem solving, research, planning, and iteration support with ChatGPT. I also use AI tools for selected personal productivity and ideation workflows when they are a good fit.

Practical AI Discipline

I use AI as a thinking partner to question assumptions, explore alternatives, and refine decisions. Human judgment remains accountable for consequential decisions; I validate outputs in the context of the systems and people they affect.

My approach to AI is pragmatic rather than hype-driven. I care about preview-first behavior, auditability, quality gates, test coverage, operational visibility, and human review. That mindset comes from having lived in QA, production support, and engineering leadership roles where “interesting” is never enough unless the software is reliable and maintainable.