Two of the most striking themes from Aaron Levie's recent conversations with enterprise IT leaders are connected in a way most people are not discussing.

The first: enterprises are moving toward a multi-model world. "Lots of interest in layers that can route workloads to different models for cost or performance reasons," Levie wrote. Companies want the flexibility to swap models in and out — Claude for complex reasoning, GPT for speed, open-weight models for cost-sensitive tasks — without rebuilding their stack every time a new model launches.

The second: talent for AI deployment "remains a major issue." Companies cannot find enough people who know how to build, manage, and maintain AI agent systems. And those who do exist are expensive, scarce, and in demand from every major enterprise on the planet.

These two problems are really one problem. And it has a straightforward solution.

The AI landscape in 2026 looks nothing like it did 18 months ago. No single model dominates every task. Anthropic's Claude handles nuanced analysis and long documents exceptionally well. OpenAI's models excel at rapid generation. Google's Gemini brings multimodal capabilities. Open-weight models like Llama offer cost advantages for routine tasks.

Smart companies are not betting on one model. They are building — or looking for — routing layers that automatically direct each task to the right model based on complexity, cost, and performance requirements.

The problem is that building a multi-model routing system from scratch requires exactly the kind of specialized AI engineering talent that Levie says nobody can find.

Viktor lives inside Slack and Microsoft Teams. You @mention it in a thread the same way you would ask a colleague. The output — a PDF, a report, a task created in your CRM, an email drafted in Gmail — lands where it should land.

Under the surface, Viktor routes tasks across multiple AI models automatically. Complex analytical work goes to the most capable model available. Routine generation tasks go to faster, more cost-effective models. The user never needs to know which model is handling their request — they just get the best result for the task.

This is the multi-model routing layer that Levie's IT leaders are trying to build. It already exists. It requires zero specialized AI talent to deploy.

Levie noted that most companies "necessarily have to train for it internally due to a shortage of talent being trained on this on the outside." That is true if you are building your own AI infrastructure. It is not true if you choose a platform that handles the infrastructure for you.

Viktor connects to over 3,200 business tools — your CRM, email, project management, cloud storage, analytics platforms. Connecting a new tool takes minutes, not months. No API development. No custom middleware. No dedicated AI team required.

The IT leaders Levie spoke with are also asking about data moats — how to capture proprietary data in formats that agents can use. Viktor solves this by working within the enterprise tools where that data already lives. It does not require data migration, new databases, or custom data pipelines. The data stays where it is. The agent goes to it.

You get $100 of free credits to begin. No time limit, no commitment. That's enough to do real work and see what Viktor can actually do before you spend a penny. There's also $50 off your first bill. You must use this exact link to receive both benefits: AIThatDelivers.com

Disclosure: Some links in this article are affiliate links. If you choose to get started with Viktor using the links provided, I may receive a commission — at no additional cost to you. I only recommend tools I use and believe in.