Three boring questions to ask before you buy AI for a staffing agency
Helge Bjorland, CEO of Globus.ai, on the three boring questions that show whether an AI agent for a staffing agency does real work or only demos well.
Helge Bjorland, CEO of Globus.ai. First shared on LinkedIn.
I am a data scientist by trade. Twenty plus years building AI models and AI software. The CEO role, I am still mostly faking.
I say it as a joke on sales calls. There is a serious point underneath it.
When a vendor tells me their platform is AI native, I do not really hear the pitch. I read the data model. Standard objects, custom objects, APIs bolted through an object layer. That is not a new model. That is a relabelled CRM with a leaf taken out of Salesforce's book and put in the press release.
You should not need a data science background to see that. But buying AI for a staffing firm right now is genuinely hard. Every product page says agents. Every demo looks magical. The difference between a language model with a nice interface and an agent that actually books shifts in your systems does not show up in a demo. It shows up three months in, when someone asks what it did last week and nobody can answer.
Ask the boring questions
So ask the boring questions instead. What does it do when nobody is prompting it? Where does it write? Which record did it touch yesterday, and can I see the log?
The founders who light up at those questions are building something. The ones who pivot back to the vision slide, well. You know.
If you are evaluating AI this autumn, send me the vendor's answer to the log question. I will tell you honestly whether it means anything.
Related reading: What is an AI agent? Explained for staffing and recruitment and EU AI Act and recruitment: what staffing agencies need to know.
Want to see what an AI agent does in a real agency, with the log to prove it? Book a conversation.
