Dallas AI Developer — Financial Services and Microsoft-Shop AI

Dallas as an AI buyer market

Dallas / Fort Worth runs on financial services, telecommunications, defense, and healthcare — and on Microsoft. The DFW corporate tech stack tilts harder toward .NET / Azure / Microsoft 365 than any major U.S. metro outside Redmond itself. JPMorgan Chase, Capital One, Charles Schwab, Texas Capital, AT&T, Lockheed Martin, Bell Helicopter, Tenet Healthcare, Children\'s Health, McKesson — the Dallas buyer pool is dominated by enterprises whose AI demand sits inside an Azure AD / Entra ID perimeter, not outside it.

The practical implication: most Dallas AI engagements need a self-hosted or Azure-tenant-scoped deployment posture, not a third-party API. The compliance posture in financial services (FFIEC, SOX, PCI), telecommunications (CPNI), and defense (CMMC, ITAR adjacency) regularly rules out OpenAI / Anthropic / Bedrock as a production path. The custom AI build is not a luxury — it is the only deployment option.

I work as the engineer who ships that build. The model is fine-tuned on the buyer\'s corpus. The inference runs in the buyer\'s Azure tenant or on-prem hardware. The eval harness, monitoring, and deploy runbook belong to the buyer. There is no third-party LLM API in the production path.

Dallas verticals I work in

• Financial services. Document extraction (loan applications, insurance correspondence), customer-service triage, claim review, internal copilots for compliance teams.
• Microsoft-shop enterprises. AI features inside the Azure AD / Entra ID perimeter, Power BI-adjacent dashboards, Microsoft Graph integrations, .NET-stack-friendly AI pipelines.
• Defense and aerospace. Subcontract-shaped engagements with prime contractors; no clearance work without prime sponsorship.
• Healthcare-adjacent. Tenet, Baylor Scott & White, Children\'s Health — scheduling, claim review, audit logging, contractor portals. Not directly clinical / EHR work.

DFW geography

Plano (Toyota, JPMorgan, Texas Health), Frisco (corporate campuses, FC Dallas, the Star District), Irving (DFW airport adjacency, ExxonMobil, McKesson), Arlington (entertainment district, Texas Rangers, Dallas Cowboys, GM assembly), Las Colinas (Pioneer Natural Resources, Verizon, Citi), downtown Dallas (Bank of America Plaza, JPMorgan, Comerica), Richardson (the Telecom Corridor — AT&T, Texas Instruments, Cisco). Each sub-metro has its own dominant employers and its own commute pattern; the engagement shape adapts to whichever sub-metro the buyer operates in.

What I build for Dallas AI buyers

• Self-hosted LLM systems. Llama 3 / Llama 4 / Mistral / Qwen fine-tuned on the buyer\'s corpus, served via llama.cpp, vLLM, or Triton inside the buyer\'s infrastructure.
• Azure-tenant-scoped deployments. Azure ML, Azure OpenAI (Microsoft-tenant scoped, not public OpenAI), Cognitive Services-adjacent pipelines.
• RAG pipelines. Vector stores (pgvector, Qdrant, Azure AI Search), citation-first answers, audit trails.
• Document-extraction workflows. Loan applications, insurance forms, telecom service records — structured data out of unstructured inputs.
• Eval + monitoring tooling. The harness that proves the model is good enough on day one; the monitoring that proves it stays good in week 26.