Kana ads

Two questions I keep mulling over in my head: - What is the mandate of a Chief AI Officer? What are these new officers being asked to accomplish? - Are the folks who accept this role being set up for failure before they even begin? One person owning AI adoption across finance, marketing, sales, and ops sounds tidy on paper. But is that fair to ask of anyone? Those are only four of the core domains, and nobody can possibly understand them all super well. That's why I think there may be issues in the standard job description for this role. To be clear, I'm not arguing the CAIO role should go away. I'm arguing it's destined to be overloaded from the get-go. In the recent video discussion that Tom and I recorded about our Agentic Divide research, I get into the first move I'd actually make as a marketer, and it starts with what you're willing to learn. So who do you think should own agentic marketing? The CMO? The chief AI officer? Or should it follow a shared ownership model? See how 225 enterprise leaders responded in Kana's Agentic Divide report https://lnkd.in/gcdphiXU, and then let me know what you think. In particular, I'd really love to hear directly from CAIOs on how they're handling this.

The LLM call is the least interesting part of the system. What's actually interesting is what surrounds it: which model do we fall back to, and what are we allowed to say? I'll use one example: a customer support triage agent. It reads a long support thread, classifies the issue, calls a tool to fetch similar past tickets, and drafts a response for the human agent to send. The default approach to keeping the triage agent "safe" is to add sentences to the system prompt: "Do not include customer SSNs or credit card numbers in the draft. Do not mention competitors. Do not give legal or medical advice." This works until the moment it matters. Prompt rules are suggestions to a probabilistic system. Guardrails are policy-enforced outside the model, on both the input and the output. The reason guardrails exist as a separate layer is that the people responsible for what the model is allowed to do are usually not the people writing the prompt. Legal, compliance, security, and ops have a legitimate interest in saying "the triage agent must never output a credit card number" without having to file a PR against your feature code. If the only way to enforce that is to edit a Python file, you have an org problem disguised as a tech problem. Cascades pay back the first time a provider has a regional outage. Guardrails pay back the first time someone outside engineering asks "how do we make sure the model can never do X?" When that question comes (from legal, from a customer's security review, from a regulator) you want the answer to be a config diff, not a sprint. So, sure, you can dismiss them and say that guardrails like these are merely infrastructure busywork. But, in truth, they are actually the difference between "we shipped an AI feature" and "we run an AI product." Read more in my Substack article here: https://lnkd.in/gXEC5q9z This is the architecture problem I set out to solve at Kana. You can learn more about what we're doing at: www.kana.ai

Revenue leaks hide at the daypart, format, and device level. Most yield teams are filling that gap manually. Automate it with Kana.

Hours lost to manual GAM checks. Then makegoods hit anyway. Kana catches under-delivery before it costs you. See how it works.

Kana gives your ad ops team visibility across your entire GAM portfolio. Most teams see savings that justify Kana's cost within one month.

Kana connects directly into your existing GAM stack. No custom builds, no IT tickets, no rip and replace. See how it works.