Sidu Ponnappa + realfast blog agent
May 20, 2025
Recently, we ran a closed-room session on AI Transformation with the IT Services leadership group of a Big Four client. This piece is a generalized version of that conversation.
It's meant for services leaders who already sense that something fundamental is shifting, and are looking for frameworks to reason through these changes.
I. The strategic imperative: Go AI-native or get left behind
We're all already selling AI Transformation to our customers - but many of us in the services industry are ignoring the very obvious Great Filter event staring us in the face.
Infosys has done some of the clearest public thinking here, making it explicit that a services firm cannot credibly lead an AI transformation unless it has itself undergone such a transformation.
AI Transformation is not a deliverable.
It's a lived capability, proved through superhuman speed and quality of service.
In other words, clients can tell the difference because it will be obvious. If you haven't done the work internally, you're not delivering an implementable strategy, you're delivering theatre.
And those who have done the work? They aren't incrementally better; they're operating on a different curve entirely. These are firms that are 10X to 100X faster, dramatically cheaper, and consistently higher in quality. They're not improving delivery, they're changing the physics of delivery itself.
The first firms to make this leap will be the ones already in the business of selling transformation.
If you're in IT Services, you're on the front lines.
II. Proof of change: The collapse is already here
Take a look around. In Q1 2025, two major service lines, illustration and research, saw pricing and timelines drop by 10-100X.
Thanks to models like GPT-4o ImageGen and the rise of Deep Research tools from OpenAI, Perplexity, and Anthropic, illustration and research deliverables that took days and cost hundreds of dollars in Q4 '24 now take minutes and cost a few cents just one quarter later.
I wrote this post 10X faster thanks to our internal AI Agent stack.
This isn't just a fluctuation, it's a structural reset. And the same pattern is sweeping through every function tied to knowledge work - task by task, skillset by skillset.
And it's accelerating.
III. The new delivery stack
Services firms used to scale by adding people. That playbook no longer applies. AI has decoupled revenue from headcount, and what replaces it is not simply cheaper execution, it's a new delivery stack altogether.
AI has decoupled revenue from headcount.
We're moving into a model defined by high gross margins and high revenue per employee. Human judgment is still involved, but it's limited to the highest-leverage decisions. Everything else gets handed off to machines.
The core mechanic of services, human reasoning, has now been commoditized. Token costs are collapsing at a pace of 10X annually, and so is the cost of producing high-quality reasoning.
As reasoning becomes productized, service delivery starts to resemble product development.
RFPs, PRDs, codebases, test suites, these can now be generated by AI at speed and scale. In 5 years, projects that once took six months will be delivered in a matter of days.
IV. Business model upgrades: What breaks and what replaces it
SaaS gets unbundled
Custom software isn't just viable, it's cheaper. Bespoke, managed software delivered by AI Native firms is already undercutting SaaS in total cost of ownership. Need a CRM or HRMS? You can have one, tailored to spec, in a matter of days. Delivered as a managed service. Priced at SaaS levels.
Time & materials is done
When most of the work is done by machines, time-based billing models break. The upside gets left on the table. We're seeing a shift toward pricing models that reflect business transactions executed above SLAs, not labor hours.
Think usage, actions, output, outcomes.
The rise of the hands-on manager
Each individual contributor becomes a manager, not of people, but of agents. The new elite is defined by cognitive stamina, discipline in review loops, and the ability to drive results through iterative delegation and feedback.
This changes who you hire. Volume hiring is gone. What you need is a senior, highly competent team that orchestrates intelligent agents at scale.
V. When all four pillars shift, it's not the same business
None of this is cosmetic. All four foundational levers of services are moving:
Talent: Who you hire, how you train them, and how you retain them changes completely
Staffing: The ratio of seniors to juniors inverts; you need experienced operators managing AI agents
Billing: Pricing models abandon effort-based logic in favor of value-based structures
Utilization: AI powered client transactions per employee replaces hours billed per employee
You're not iterating the old model. You're building a new one. The AI Native services firm operates with software economics layered over expert judgment.
Final thought: The U-turn is coming
The bigger and more successful your existing business, the harder this gets. This is the Innovator's Dilemma, live and unfiltered. You'll cannibalize revenue. You'll put sacred P&Ls at risk. But the headcount pivot isn't optional, it's inevitable.
The firms that get ahead of it? They won't just stay relevant. They'll define the new category.
At realfast, we don't bolt AI on, we redesign around it. We optimize for leverage over labor, and we're always up for working with those ready to do the same.
Start your AI transformation journey today
Ready to unlock the true potential of AI in your services business? Schedule a zero-cost workshop with our team to identify immediate high-impact opportunities. We'll help you deploy AI agents that deliver measurable business outcomes in days, not weeks, cutting implementation times in half while maintaining quality.
