AI Software Architect

AI Software Architect
Hybrid / Dallas, TX (2-3 days Onsite)
Perm Role

Our client is building the AI-native engineering organization of the future — one where small, durable teams direct AI to deliver software faster and more reliably than traditional models allow.

This role sits inside that effort from day one: working with the latest in agentic AI development, helping shape the practices and platforms the next generation of technology delivery will be built on, and contributing directly to a capability that compounds in value with every project completed.

The Sr Architect owns the technical shape of what the pod builds and the guardrails insidewhich AI operates. They set architecture, encode the Client’s Enterprise Architecture and API Modernization Standards into forms the pod and its AI tooling apply automatically, and are the accountable author of Architecture Decision Records when a deviation is genuinely warranted.

In an AI-Native pod, the guardrails are the product as much as the code: a well-specified ARCHITECTURE.md, a precise CLAUDE.md, and deterministic permission boundaries are what allow AI to move fast without drifting off-standard.

Beyond setting the guardrails, the Sr Architect is accountable for ensuring every member of the pod team works within the client’s standards day to day — and for capturing what each project reveals about those standards so that lessons learned flow back to improve them for every team that follows.

Responsibilities:

• Set system architecture — modular-monolith boundaries, integration patterns, data flows, and the mandatory lakehouse and observability wiring.
• Author and maintain ARCHITECTURE.md — the system design document that doubles as the DDP High-Level Tech/Security Design and the primary reference for every AI coding session.
• Author and approve ADRs — held to the “what would have to be true” exception bar; push back on convenience dressed up as justification.
• Interface with the Architecture Committee — on open decisions and carry committee direction into the pod.
• Codify standards as guardrails — turning EA Standard rules into CLAUDE.md conventions, review checklists, and settings.json deny rules rather than relying on memory or manual review.
• Ensure APIM-fronted integration — OAuth 2.0 / OIDC identity, Terraform-managed infrastructure, and OpenTelemetry instrumentation are non-negotiables in the pod’s output.
• Review high-risk changes — for standards compliance and architectural integrity before they reach production.
• Maintain EA and API Standard compliance — for everything the pod ships; escalate exceptions to the Architecture Committee with a supporting ADR.
• Contribute to the shared standards skills — caliber-ea-standard and caliber-apistandard via Skills Library Updates at project close.
• Evaluate and adopt new patterns — as the AI platform and tooling landscape evolves, in alignment with open ADRs (ADR-011 through ADR-014).
• Hold the pod to the client’s standards — actively coach pod members on EA and API Standard requirements day to day, catch drift before it reaches review, and make standards a living practice rather than a document people read once.
• Lead the lessons-learned cycle — at the close of every project, own the Skills Library Updates in LESSONS_LEARNED.md so what the pod discovers — gaps, ambiguities, new patterns, anti-patterns — is formally proposed back to the Architecture Committee and incorporated into the standards.

Qualifications:

• 8+ years in software engineering with substantial architecture ownership on production systems.
• Deep command of modern application stacks: .NET 8 and/or TypeScript, EF Core / Drizzle, PostgreSQL, containerized cloud infrastructure, Terraform, and APIM-style API
• Fluency with modular-monolith design, API modernization patterns (OpenAPI 3.1, standard response envelope, facade-first migration), and secure SDLC.
• Ability to write standards with sufficient precision that both engineers and AI tooling apply them consistently.
• Sound judgment on when a deviation is warranted and the discipline to document it as an ADR.
• Bachelor’s degree in computer science, engineering, or a related field, or equivalent

Preferred Qualifications:

• Experience defining guardrails for agentic AI development (skills libraries, context files, permission models).
• Background in legacy modernization and strangler-fig migrations — directly relevant to CARE platform decomposition.
• Familiarity with lakehouse integration patterns (Iceberg / Snowflake / Databricks) and event-driven messaging.
• Cloud-agnostic architecture experience, relevant to open cloud-strategy decisions across AWS and Azure.

Core Competencies:

• Technical judgment — makes sound, pragmatic architecture decisions and knows when to standardize versus enable autonomy.
• Precision in communication — writes standards and guardrails that are specific enough to be enforced by both humans and AI.
• Principled flexibility — distinguishes real exceptions from inconvenient standards and documents the difference.
• Collaborative governance — works within the Architecture Committee model and brings decisions forward rather than making them unilaterally.
• Forward-looking — anticipates how AI tooling and the platform landscape will evolve and designs for it.