3 Big Questions Before Signing That Hot Startup Contract

I recently had a enjoyable and wide-ranging conversation with Beth Kormanik of CIO Magazine about what I’m seeing from the venture side of enterprise AI. As a seed-stage lead investor reviewing over 1,400 enterprise AI startups in the past year alone, I have a wide lens on how enterprise CIOs are responding to AI startup value propositions.

That mandate is evolving fast. In a tech forward world where AI may become your primary customer interface, you are now accountable for the core customer brand promise and experience. We are watching the collapse of product and technology into a single accountable owner. This means the CIO role is no longer primarily infrastructure. This leads to three big questions every CIO must answer before partnering with an AI startup.

1. Are You Buying Breadth — or Depth?

Most CIOs have been expected to evaluate "AI vendors" on features, like security, or productivity. That is no longer sufficient.

The more important question is strategic, do you want this because you need more organizational breadth (as in benchmarks and extended capacities) or do you need depth, as in vertical expertise?

If your organization is performing in the back quartile in a function, or behind the mean in your industry, AI with broad benchmarking data can lift you. It gives you exposure to industry best practices you don’t yet possess. But if you are already top quartile? If you are a top performing enterprise in a certain area, and you bolt on a broad AI in that area, it will mediate you and bring you down.

Average AI applied to exceptional operations produces regression to the mean. So this breadth question has bite.

AI With Depth

Updraft is an example of an AI with depth in Valor's portfolio. It is built specifically for accounting advisory firms and engages on complex technical accounting questions. It enhances specialist performance rather than diluting it. It is not broad “AI for finance.”

For CIOs looking to skill up their strength in deep areas of expertise, domain-native vertical AI that builds a tailwind in a core capability is critical—and then it allows agentic communication to distribute that insight across the enterprise. Tim LaBelle, CEO at Updraft, shared a cautionary story that illustrates what happens with an organization fails to build depth.

"Kyndryl's stock dropped 55% in a single day last week," he related. Four disclosures landed before the market opened:
1️⃣ the company couldn't file its 10-Q on time
2️⃣ material weaknesses were expected across multiple reporting periods,
3️⃣ the CFO and General Counsel departed immediately
4️⃣ and the SEC's Division of Enforcement had already been requesting documents.

"$3 billion in market cap was gone before lunch. When investors can no longer rely on the numbers, everything gets repriced — fast. That's what a trust event looks like," says Tim.

The accounting skill shortage is getting worse every year. Proper financial reporting depends entirely on having the right people, doing the right work, with the right judgment. And that's becoming harder to guarantee.
A material weakness doesn't appear overnight. It accumulates quietly in the gaps between what the work requires and the capacity of the team doing it. An AI, like Updraft, can add depth and capacity to the team you have.

For CIOs contemplating a new AI like an Updraft, the core question to ask yourself is, are we about to upgrade our performance in a niche functional area that is not our core? If so, that's a great AI for you.

Ask yourself, where are we average? This is where you want to partner with AI startup leaders like Tim.

2. What Is Your Core Customer Promise — and Will This AI Strengthen or Weaken It?

This is the most under-asked critical question in enterprise IT.

Why are customers choosing you? What is your core value, core customer promise, brand promise? Where are we truly exceptional / best in class?

Historically, we didn’t require CIOs to own this. That era is over.

If AI becomes the layer through which your product is experienced, then build-versus-buy decisions are now brand decisions. The number one thing that you are best in the world at and you intend to lead the globe in—that needs to be under your remit as technical and product leadership. But everything else--find the agent for it.

AI With Breadth

So what does it look like to add on AI with to extend your breadth? Valor portfolio firm Sailes is a good example of breadth deployed properly at customers like Schneider Electric. It enhances sales organizations by systematizing and scaling high-performing behaviors. It does not replace core brand value; it accelerates it. Sales as a function is not the core brand promise for most enterprises. With a sales-leading AI like Sailes, which comes with ROI guarantees, you gain the benchmarked best practices at a scale beyond your own organization--but still with clear ownership of your own process and data. That's where a broad, functional AI can support you in staying focussed on what you do best.

3. Are You Using AI to Prepare for the Disaster You Never Modeled?

Most CIOs think of AI as productivity tooling. That is perhaps the least interesting application, although an important one to master.

One of the most powerful uses of AI is resilience modeling and recovery. Imagine the worst thing that you could actually imagine happening to your organization. With synthetic data paired with your own data, you can actually run faster-than-real-world simulations. This changes the resilience calculus.

Example: Autonoma — AI for Real World Simulation

Autonoma, an Auburn University spinout, can simulate a 500-year flood on existing infrastructure, a tornado on your headquarters building, a catastrophic ice storm closing your data center—something that historically would have required years and perhaps millions of dollars for top consultants to model.

This is not generative AI. This is vector-based modeling and scenario intelligence. For CIOs, this is a new opportunity to future proof.

• What scenario have you never prepped for because it felt unlikely?

• What systemic failure would you only discover during crisis?

• What if you could simulate it tomorrow?

AI is not just operational lift and productivity. It is strategic foresight and resilience.

The New CIO: Infrastructure + Product + Brand Guardian

Today’s CIO is not a cost center. You are:

• Architect of competitive advantage

• Owner of resilience modeling

• Co-steward of product identity

• Arbiter of build versus buy

AI is accelerating the collapse of silos. The question is whether you will step into that expanded remit—or let it default elsewhere. Before you sign your next AI startup contract, ask:

1. Are we buying breadth or depth?

2. Does this strengthen our core customer promise?

3. Does this increase resilience—or just automate convenience?

The answers will determine whether AI becomes your moat—or your mediocrity.