Job Posting Autopsy — Week 5: Hallmark Director of Architecture for AI | Brian Curry

Series: Weekly Data Science Job Posting Review Company: Hallmark | Posted Title: Director of Architecture for Artificial Intelligence | Diagnosed Title: Director of Enterprise Architecture (AI-Focused) — legitimate but overstuffed

Overview

Hallmark — yes, the greeting card company — is hiring a Director of Architecture for AI at its Kansas City headquarters. The role sits within the Enterprise Architecture team and is responsible for AI strategy, governance, platform engineering, multi-agent system design, and cross-functional leadership. The posting requires 12+ years of IT experience, 10+ years in enterprise architecture or AI/ML systems design, TOGAF or Zachman certification, and a cloud AI/ML certification.

No salary is posted. The preferred qualifications section is longer than most job postings in their entirety.

This is a genuinely interesting role buried under a requirements avalanche. Let’s unpack it.

1. Role Identity Test

I mapped each stated responsibility to standard role archetypes: Enterprise Architect, AI/ML Architect, AI Governance Lead, Platform Engineer, and People Manager.

Define and evolve the AI reference architecture → Enterprise Architect (AI-specialized). Core EA function.
Collaborate with executive leadership on AI opportunities → Strategic Advisor / Director. Executive-facing strategy work.
Lead the Hallmark AI Accelerator Program → AI Program Director. Innovation program management.
Partner with data architects, engineers, legal, privacy, security → Enterprise Architect. Cross-functional governance.
Implement AI governance frameworks, risk levels, decision rights → AI Governance Lead. Policy and compliance.
Stand up and evolve AI platforms, sandbox environments, vendor selection → Platform Engineer / AI Infrastructure. Hands-on platform work.
Design architecture for AI agents, multi-agent systems, autonomous workflows → AI/ML Architect. Genuine AI architecture.
Establish standards for agent orchestration, memory management, tool integration, human-in-the-loop → AI/ML Architect. Cutting-edge agentic AI design.
Mentor architects and engineers, promote AI literacy across enterprise → People Leader / Evangelist. Organizational development.
AI/ML systems design: model training, inference pipelines, enterprise integration → AI/ML Architect. Hands-on technical.
Cloud and infrastructure: Azure ML, SageMaker, Vertex AI, GPU/TPU, Kubernetes → Cloud/AI Infrastructure. Platform engineering.
Security: prompt injection, model extraction, NIST AI RMF → AI Security Specialist. A niche specialty.
Partner with data governance to align AI with enterprise data platforms → Enterprise Architect. Data architecture alignment.

Estimated role composition:

Enterprise Architecture / Strategy: ~30%
AI/ML Architecture (genuine): ~25%
Governance / Compliance / Security: ~20%
Platform Engineering / Infrastructure: ~15%
People Leadership / Evangelism: ~10%

Diagnosis: Unlike the previous “AI” posting in this series (the Dynamics 365 role masquerading as AI), this one is legitimate. The responsibilities describe genuine AI architecture work — agentic systems, multi-agent orchestration, LLM security, inference pipelines. The title is honest. The scope, however, is massive. This is an enterprise architecture director, an AI technical architect, a governance officer, a platform engineer, a security specialist, and an organizational evangelist rolled into one.

2. Skill Breadth Score

The posting’s preferred qualifications section is extraordinary. Counting across basic and preferred:

Required:

Enterprise architecture (10+ years)
TOGAF or Zachman framework certification
Cloud AI/ML certification (Azure AI Engineer, AWS AI Practitioner)
AI/ML systems design (model training, inference, integration)
Cloud platforms (Azure ML, SageMaker, Vertex AI)
GPU/TPU infrastructure and Kubernetes/Docker
AI security (prompt injection, model extraction, NIST AI RMF)
Leadership (3+ years)

Preferred (abbreviated — the full list is much longer): 9. GDPR, ISO 27001/27002/20000–1/42001 compliance 10. AI fairness, transparency, and explainability 11. AI governance board leadership 12. AI center of excellence program management 13. TensorFlow, PyTorch, Hugging Face, OpenAI, Llama, Grok, Anthropic 14. Azure AI Foundry, AWS Bedrock, H2O.ai 15. API/A2A/MCP connections for LLM integrations 16. NIST Cybersecurity Framework 17. Cloud governance 18. Python, Java, R, Golang 19. SAP LeanIX or EA tooling 20. Snowflake, Informatica, Power BI, Teradata 21. Office 365 & Power Platform (SharePoint, Visio, Power Automate) 22. Multi-year architecture roadmap development 23. Business, data, application, and technology layer modeling 24. Data science, statistics, algorithm design

Total: 24 distinct skill areas.

Benchmark: Even for a director-level architecture role — which legitimately requires breadth — 24 skill areas across enterprise architecture, AI/ML engineering, security compliance, cloud infrastructure, data engineering, programming, governance frameworks, and office productivity tools is excessive. The posting wants a person who is simultaneously a TOGAF-certified enterprise architect, a hands-on ML engineer, an AI security specialist, a governance policy author, fluent in four programming languages, certified in three cloud platforms, and experienced with Snowflake, Teradata, Power BI, SharePoint, and Visio.

The inclusion of “Office 365 & Power Platform (inc. SharePoint, Visio, Power Automate)” in the same preferred qualifications list as “GPU/TPU infrastructure” and “prompt injection defense” is a tone-setting detail.

Breadth score: 24/8 — exceeds focused-role threshold by 200%.

3. Red Flag Inventory

Flag 1: Title–Responsibility Mismatch — NOT PRESENT

The title “Director of Architecture for AI” accurately describes the role. This is a director-level position responsible for AI architecture across the enterprise. The responsibilities align with the title. Credit where it’s due.

Flag 2: Unicorn Posting — CONFIRMED (SEVERE)

The combined required and preferred qualifications describe not one person but an entire AI center of excellence team. Enterprise architecture certification + hands-on ML engineering + AI security expertise + governance framework authorship + cloud platform mastery across three vendors + four programming languages + data engineering tools + office productivity tools. Nobody has all of this. The posting is aspirational fiction dressed as a job description.

Flag 3: The “Name Every AI Company” Problem — NOTED

The preferred qualifications include: “TensorFlow, PyTorch, Hugging Face, OpenAI, CoPilot, SageMaker, Llama, Grok, Anthropic.” This mixes ML frameworks (TensorFlow, PyTorch), model libraries (Hugging Face), API providers (OpenAI, Anthropic), open-source models (Llama), a social media AI product (Grok), a cloud ML platform (SageMaker), and a Microsoft productivity copilot (CoPilot) into a single bullet point as if they’re interchangeable. They are not. This reads like someone Googled “AI tools 2025” and pasted the results into a qualifications list. It suggests the hiring team doesn’t have deep technical fluency in the space they’re hiring for.

Flag 4: No Salary Range Posted — RED FLAG

Another KC-area posting with no salary transparency. For a director-level role requiring 12+ years of experience and multiple certifications, this is a significant omission. Director of Enterprise Architecture roles nationally pay $205K–$354K total on Glassdoor. AI Director roles average $302K nationally. Without a posted range, candidates are negotiating blind against a company that knows exactly what its budget is.

Flag 5: 12+ Years Experience for an Emerging Field — CONTRADICTION

The posting requires “at least 10 years’ experience in enterprise architecture or AI/ML systems design.” AI/ML systems design at enterprise scale barely existed as a distinct discipline 10 years ago. The LLM and agentic AI work described in this posting (multi-agent orchestration, RAG architectures, prompt injection defense, MCP connections) has only been relevant for 2–3 years. What the posting actually needs is deep enterprise architecture experience plus recent AI/ML expertise — but the way it’s written implies someone who has been doing AI architecture for a decade, which filters out candidates who transitioned from EA to AI more recently and may be better suited.

Flag 6: Private Company Dynamics — NOTED

Hallmark is a private company (owned by the Hall family). This means no public stock, no RSU grants, no equity transparency. Compensation is base + annual incentive plan + 401K match + profit sharing. The profit sharing is a meaningful benefit at a profitable private company, but candidates coming from public tech companies should understand that total comp will look structurally different — more cash-heavy, less equity upside.

4. What’s Missing vs. What’s Overrepresented

Present and expected for a Director of AI Architecture: AI strategy, reference architecture, governance, cross-functional leadership, platform decisions, mentorship, security awareness, cloud platform knowledge.

Missing: Any mention of budget ownership, team size, or direct reports — unusual for a director-level role. Also absent: specific business use cases that ground the AI work in Hallmark’s actual product domain (media, retail, greeting cards, e-commerce).

Overrepresented: The preferred qualifications list, which tries to cover every AI tool, framework, cloud platform, compliance standard, programming language, and data tool that exists. Also overrepresented: governance and compliance language relative to actual building-and-shipping.

5. Compensation Analysis

Posted range: Not disclosed. Annual incentive plan, merit increases, 401K with match, and profit sharing mentioned.

Market Benchmarks (2026)

Director of Enterprise Architecture (national):

Glassdoor: $213K–$354K total (25th–75th percentile), average ~$272K
Salary.com: ~$205K average base

AI Director (national):

Glassdoor: $231K–$402K total (25th–75th percentile), average ~$302K (limited data)

AI Architect (national):

Glassdoor: $139K–$252K total (25th–75th percentile), average ~$185K
Salary.com: ~$180K average base
Robert Half: $143K–$197K starting range

Hallmark context:

Glassdoor: VP-level roles at Hallmark top out around $244K. Engineering department averages ~$144K (Comparably).
Hallmark is a private, KC-based consumer goods company — not a tech company. Comp will reflect this.

Analysis

This is where it gets uncomfortable. The national market for a Director of Enterprise Architecture with AI specialization sits at $200K–$350K+ total. But Hallmark is not Google, not even Cerner. It’s a private consumer products company in Kansas City with an engineering department that averages ~$144K.

Based on Hallmark’s comp structure and the KC market, this role likely pays in the $160K–$220K base range, with total comp (incentive plan + profit sharing + 401K match) potentially reaching $190K–$260K. That would be competitive for a director role at Hallmark in Kansas City — but significantly below what this skill set commands at companies where AI is a core product rather than an internal transformation initiative.

The trade-off is real: a candidate taking this role is betting on the appeal of building an AI practice from scratch at a beloved American brand, with likely better work-life balance than a big tech role, in exchange for $50K–$150K less in total compensation than the same skill set would earn at a tech company. That’s a valid personal choice — but candidates should make it with eyes open.

Comp verdict: Likely competitive for Hallmark and KC, significantly below market for the skill set nationally. The missing salary range makes precise evaluation impossible, which continues to be a pattern with KC employers and is worth calling out. The profit sharing and private company stability are real benefits that partially offset the base gap.

6. The Verdict

What Hallmark titled this: Director of Architecture for Artificial Intelligence

What this actually is: A Director of Enterprise Architecture with a genuine AI mandate — and a qualification list that describes an entire department. The title is honest. The scope is not.

Who this role is actually for: A senior enterprise architect (10+ years, TOGAF certified) who has recently developed deep AI/ML expertise — particularly in agentic systems, LLM integration, and AI governance — and wants to build an AI practice at a major consumer brand. This person exists, but they are rare, and they will need to ignore 60% of the preferred qualifications to apply with confidence. If you’re an AI architect from tech who wants to slow down, own a domain, and build something from the ground up in a lower-cost-of-living market, this could be genuinely appealing.

Bottom line: Hallmark gets credit for an honest title and a genuinely interesting charter — how many people get to define AI strategy for a 115-year-old American institution? But the posting undermines itself with a qualifications list that reads like a keyword dump. The “name every AI company” bullet point (OpenAI, Grok, Llama, Anthropic, CoPilot in one breath) suggests the hiring team is still learning the landscape, which is fine — that’s exactly why they’re hiring this role — but it means the successful candidate will need to educate upward as much as execute downward. That’s either exciting or exhausting depending on where you are in your career.

Job Posting Autopsy is a weekly series that applies structured analysis to data science job listings. The goal is not to mock employers but to help candidates make informed decisions — and to push the industry toward more honest role descriptions. Have a posting you’d like reviewed? Drop it in the comments.

Job Posting Autopsy — Week 5: Hallmark Director of Architecture for AI