Chief AI Officer Advisor — Business

Chief AI Officer Advisor is a strategic advisory skill for startup CAIOs and founders, covering four key decisions: model build-vs-buy (API vs fine-tune vs in-house), AI risk classification under EU AI Act and US state laws, API-to-self-hosted cost economics with breakeven analysis, and AI team org evolution. It provides decision frameworks, Python calculators, and governance checklists to guide AI strategy without hype.

c-level-advisor claude-code

📂 View skill on GitHub ↗ 🚀 How to use

✨ What it does

Model build-vs-buy calculator with 3-year TCO analysis.
AI risk classifier for EU AI Act tiers and US state law obligations.
API-to-self-hosted cost economics with breakeven point and sensitivity analysis.
Stage-to-role map for AI team hiring (AI engineer vs ML engineer vs research scientist).
Output standards with bottom line, decision, evidence, and action steps.
Cross-references to adjacent C-level advisors for budget, legal, and security checks.

🎯 When to use it

Deciding whether to use an API, fine-tune a model, or build in-house for a specific use case.
Classifying an AI use case for regulatory risk under EU AI Act, NIST AI RMF, or US state laws.
Calculating when self-hosting inference becomes more cost-effective than API calls.
Determining the next AI role to hire based on company stage and capabilities needed.
When the user mentions CAIO, AI strategy, model selection, foundation model, fine-tuning, EU AI Act, AI governance, or AI economics.

🚀 How to use

Trigger: /cs:freeze

Trigger by asking for AI strategy, model selection, risk classification, cost economics, or AI team planning. Provide a use case description or a JSON file with relevant parameters (volume, latency, accuracy, budget, geography, sector). Run the provided Python scripts for quantitative analysis. Example prompts:

Should we use an API or fine-tune a model for our customer support chatbot?
Classify our AI hiring tool under the EU AI Act and list required controls.
At what token volume does self-hosting beat the API for our workload?

📄 Output: A decision brief with bottom line, evidence, and action steps, plus optional calculator outputs and governance checklists.

📦 Add this skill to Claude Code

# 1. Get the skills repo
git clone --depth 1 https://github.com/alirezarezvani/claude-skills /tmp/claude-skills

# 2. Copy this skill into your project (or ~/.claude/skills for all projects)
mkdir -p .claude/skills
cp -r /tmp/claude-skills/c-level-advisor/skills/chief-ai-officer-advisor .claude/skills/chief-ai-officer-advisor

Skill source: c-level-advisor/skills/chief-ai-officer-advisor/SKILL.md

⚠️ Good to know

Strategic only; does not cover tactical AI/ML engineering (RAG, agents, prompt engineering, evals, deployment).

❓ FAQ

What is the default path for model selection?

The default path is to use a frontier API, as frontier models are 10-100x more capable than most teams can fine-tune in-house, and it is cost-effective below ~100M tokens/month.

When should we consider fine-tuning a smaller model?

Fine-tune when you need domain-specific behavior that can't be prompted into an API, high volume reduces API cost, latency budget is under 500ms, or you need specific style/format consistency.

What is the typical breakeven point for self-hosting vs API?

The typical breakeven is 100M-500M tokens per month, depending on model size and quality tradeoff; below this, API wins.

🤖 Overview, features, install steps and FAQ were generated from the project's SKILL.md on Sep 4, 2026. Always check the original source before running commands.