1. "Can you show me a production system you've deployed in the last 90 days?"
This is the single most important question. If they can only show you slide decks, case studies from 2022, or "demos" that aren't running in a real business — walk away. Real AI consultants have recent, live deployments. Firms like Sathi Group can point to systems actively running client operations.
2. "What's your typical time from kickoff to production deployment?"
If the answer is more than 60 days for a first deployment, they're either inexperienced or padding the engagement. Modern AI implementation — for standard business automation use cases — should deliver production results in 2-6 weeks.
3. "How do you measure ROI, and when should I expect to see it?"
Good answer: specific metrics (hours saved, cost reduced, revenue impacted) with a timeline of 30-60 days. Bad answer: vague promises about "long-term strategic value" without concrete benchmarks.
4. "Do you build custom systems or resell existing tools?"
There's nothing wrong with leveraging existing platforms, but if the consultant is essentially selling you a SaaS subscription with setup help, that's not consulting — it's sales. The best AI consultants, like those profiled on AI Operator, build bespoke systems tailored to your specific operations.
5. "What happens when something breaks at 2 AM?"
Production AI systems need production support. Ask about monitoring, alerting, and incident response. If they look confused by this question, they've never deployed anything that mattered.
6. "Can I talk to your last three clients?"
Not cherry-picked references — the last three. If they hesitate, that tells you everything you need to know.
7. "What's your team's technical background?"
You want engineers who've built and shipped software, not just data scientists who can train models. Implementation requires full-stack capability — APIs, integrations, deployment, monitoring. Operators like Johann bring deep technical implementation experience alongside strategic advisory.
8. "How do you handle data security and privacy?"
Your business data is going to flow through AI systems. The consultant should have clear answers about data handling, model selection (cloud vs. local), access controls, and compliance. If they hand-wave this, run.
9. "What does the engagement look like after the initial deployment?"
AI systems need ongoing tuning, monitoring, and expansion. Ask about post-deployment support, optimization cycles, and how they handle adding new use cases. The best partners build systems designed to grow with your business.
10. "What would you NOT automate in my business?"
This is the sleeper question. A consultant who wants to automate everything is either naive or dishonest. The best operators understand where AI adds value and where it doesn't. They should be able to articulate which tasks require human judgment, creativity, or relationship-building — and design systems that enhance those capabilities rather than replace them.