Board and customers are demanding AI features. Your engineering team is still lean from the 2023–24 cuts. Headcount approval isn't coming. Here's what's actually working.
Context: Ailoitte builds AI engineering teams for companies in exactly this situation. Sharing what we see working across multiple teams.
Why Is This Problem So Common Right Now?
The 2023–24 tech layoffs were widespread. Engineering teams at Series B–D SaaS companies cut 30–50% of headcount.
Fast-forward to 2025: competitors who didn't cut as deep — or who maintained AI engineering investment — have shipped features that are now visible to customers.
The board wants AI on the roadmap. Customers are asking. And the engineering team is still operating lean with a headcount freeze still politically in effect.
Why Rebuilding the Team Isn't the Right First Move
For scoped AI feature delivery, rebuilding headcount has three problems:
Timeline. 6–12 months to hire, onboard, and ramp a senior AI engineer. The board's Q3 roadmap expectation and the hiring timeline don't overlap.
Political difficulty. Adding headcount that was just cut is a hard internal conversation. The CFO who approved the cuts is still in the room.
Cost. $200K–$300K annually per AI engineer vs. $30K–$60K for a scoped engagement with a fixed outcome.
What Does an "AI Engineering Arm" Actually Mean?
An AI engineering arm is an external team that operates like an internal one. Not a consultancy. Not a separate product track.
Engineers embedded in your sprint, working in your codebase, attending your standups, shipping to your production environment.
The difference from a typical agency:
- Code ships to your production, not a demo environment
- IP stays 100% with your company
- Engineers are assigned specifically to your team, not rotated across accounts
- Start time: 2 weeks, not 6 months
How Does the Budget Conversation Change?
The companies moving fastest on this have reframed the internal approval conversation.
❌ "We need to hire AI engineers" → stalls at headcount approval.
✅ "We need to ship this AI feature — here's the project budget" → routes through a different approval path.
$30K–$60K as a project budget vs. $200K–$300K annually as a hire are evaluated differently by most finance teams — even when the output is equivalent.
What Gets Shipped?
Typical engagements for post-layoff AI delivery:
- AI search or assistant layer on an existing SaaS product
- RAG-based knowledge system for internal or customer-facing use
- AI feature sprint matching a competitor's recently shipped capability
- Automation of a specific manual workflow inside the product
The Reframe That Unlocks Movement
The engineering leads making progress right now aren't winning the headcount argument. They're bypassing it — by reframing AI delivery as a project budget conversation instead of a hiring conversation.
The output is the same. The approval path is completely different.
Ailoitte's AI Velocity Pods are structured specifically for this model.
What does your internal conversation around AI headcount look like right now? Drop it in the comments — curious what constraints others are navigating.
Top comments (0)