How I Work
Every engagement is different — but the shape of the work isn't.
Getting AI right across the wider customer support function isn't about applying a framework. It's about understanding how your operation actually runs, deciding where AI genuinely belongs and where it doesn't, and making sure the foundations are in place for it to deliver.
This page describes the typical shape of an engagement — what we'll do together, how long it takes, and what good looks like along the way. It's a guide, not a template. The specifics adapt to your context.
01 - The Engagement
A typical engagement runs 8-12 weeks and moves through four phases.
Phase 1 - Diagnose (Weeks 1-2)
The first two weeks are about understanding your customer support operation as it actually runs — not as the org chart describes it.
I look at demand drivers (what's generating the tickets, and why), cost-to-serve patterns (where time and money go, and what's driving that), knowledge architecture (what your agents can find quickly, and what they can't), escalation paths (where issues go when the first line can't solve them, and how well that works), churn signals (what customer behaviour tells you about risk), and voice-of-customer loops (how customer feedback influences product and strategy — or doesn't).
If you've already deployed AI, I assess what's running, what it's actually doing, and how it's integrating with the wider operation.
The output of this phase is a clear, written picture of where your customer support operation is working, where it isn't, and where the real opportunities for AI are.
Phase 2 - Design (Weeks 3-6)
With the diagnostic in hand, we move into design. This is where 20 years of operational leadership does its work.
We look at what AI should handle, what humans should handle, and how they interact. We design or redesign the foundations that determine whether AI can genuinely deliver — knowledge architecture, escalation paths, voice-of-customer loops, team structure, and how AI fits into all of it.
The aim isn't an AI-first operation. It's a well-run customer support function where AI contributes to the outcomes the business cares about — cost-to-serve, churn, CSAT, and the customer value that underpins all three.
Phase 3 - Deploy (Weeks 5-10)
Deployment runs in parallel with the later stages of design.
If you've already invested in AI tools, we assess whether they're the right fit for the redesigned operation — or whether the investment is better directed elsewhere. If you're deploying for the first time, we make informed choices about tools, sequence, and scope — choices you can defend to your board.
I work alongside your existing team and tooling, not around them. The goal is business outcomes, not technology outcomes.
Phase 4 - Outcomes (Weeks 10-12)
The final phase is about measurement and what's next.
Every engagement ends with a clear picture of what's changed — across the operation and its metrics — what needs ongoing attention to sustain, and a view of what the next stage of transformation looks like, should you choose to take it.
02 - What a typical week looks like
I work alongside your team, not above them.
A typical week includes one or two working sessions with your senior customer support leader, one session with your wider support team to understand operational reality and test ideas, and direct work on the deliverables — diagnostics, designs, deployment plans, measurement frameworks.
Between sessions, I'm available by email, chat, or phone for the questions that come up between formal meetings. I am not a consultant who disappears between workshops and reappears with a deck. It's closer to having a senior operational leader embedded in your team for the duration.
The commitment from your side is typically 2-4 hours per week of your senior support leader's time, and less than that across the wider team over the course of the engagement.
03 - Commercials and fit
Design-partner engagements this quarter
I'm currently taking on three design-partner engagements. This format suits the current stage of the business — deeper access, closer working relationships, and reduced fees in exchange for case-study rights and a reference agreement. First-come, first-fit.
Who this is for
Customer support, CS, or CX leaders at VP or Director level in mid-market B2B SaaS companies. Whether you're planning your first AI deployment in customer support or reassessing one already in place, the work is about making AI succeed across the wider function — not just on the deflection dashboard.
Who this isn't for
Businesses looking for a pure AI implementation consultant to pick and deploy tools. That's not where the real work lives.
Businesses looking for a generic customer support audit unconnected to their AI strategy. There are capable firms that do this work well; this isn't one of them.
Businesses at the enterprise end of the market. My approach is built for the mid-market, where the operational complexity is real but the decision-making still fits into a single senior team.
Take the next step.
If this sounds like the shape of engagement you need — whether you've got a defined brief or a rough sense that something isn't working — let's talk.
The first conversation is 30 minutes. No pitch, no deck. Just a direct exchange on what you're facing and whether this fits. I'll tell you honestly if I don't think I'm the right person for the work.
Prefer email? Write to hello@cscloud.uk directly.