Essay

Are Project Managers Still Needed? Why AI Is Becoming the Orchestration Layer

July 1, 20266 min read

Short answer: the coordination core of project management — matching resources to needs, sequencing work, routing information — is increasingly something AI does better and at larger scale than a human. What stays human is judgment: exceptions, trade-offs, trust, accountability. Project managers aren't vanishing, but their orchestration role is being automated — and that has bigger implications for how companies, and especially startups, get work done.

The revelation from a room full of organizational researchers

I sat in on a workshop my advisor ran with a group of HR and organizational-design researchers — people whose entire field is studying how organizations are structured and how they actually operate. The biggest takeaway from the room was blunt: project managers, in the traditional sense, aren't really needed anymore, because AI is better at orchestrating resources than a human coordinator is.

What struck me was that this is exactly what I've been doing with Ordana for years — having an AI orchestrate a layer of resources. It was strange to hear the academic field arrive at the thing we'd already built a product around.

Why orchestration is an AI-shaped problem

Strip project management down and a huge part of it is a matching-and-routing problem at scale: who has what, who needs what, in what order, and how information should flow. A human holds maybe a few dozen of those variables at once. An AI holds thousands and keeps updating them continuously.

That's the whole argument. When the core task is large-scale matching and routing, the tool that can hold more state and update faster wins — and that's an AI, not a single coordinating manager.

What stays human

This isn't "project managers are useless." The parts of the role that survive are the parts that were never really about coordination: handling the messy exceptions, negotiating trade-offs between people, owning relationships, and carrying accountability when something goes wrong. AI orchestrates; humans judge. The role shifts from running the machine to handling what the machine can't.

From orchestrating one company to orchestrating many

Here's the move most people miss. If AI can orchestrate resources inside a company, it can orchestrate them between companies too — and that's where it gets powerful for startups. Instead of one PM coordinating an internal team, an AI orchestration layer sits between separate startups, matching complementary capabilities and routing them to where they're needed.

That's what Ordana is: an orchestration layer between startups. Its AI surfaces collaborations across nearly 100 startups with an estimated $475K of total resources, and structures each one as an automated revenue-share deal. A founder gets the coordination output of a project manager without hiring one — and keeps their own time for the judgment calls.

The takeaway

The question isn't really "are project managers needed?" It's "what should still be done by a human, and what should be orchestrated by AI?" The coordination layer is going to AI. For startups, the highest-leverage version of that layer doesn't just organize your own team — it connects you to everyone else's resources too.