Two weeks to find out what can run without you

If every approval, every figure and every follow-up still lands on your desk, adding AI tools will not change much. First, you need to know where your time goes. The audit answers that in two weeks, with figures and a clear sequence.

I asked myself the same question first. In four months, I built 84 agents and 13 applications on my own, and they now back me up every day as a CEO. The audit starts from there: what in your week no longer needs to go through you?

Is it for you?

The audit is designed for owners, CEOs and leadership team members (managing director, CFO, sales director, COO) of SMEs and mid-sized companies with 20 employees or more, family businesses included. It is useful if:

• decisions, figures and follow-ups all wait for you;
• your teams re-key the same data into several systems;
• a few people use AI tools on their own, with no visible effect on the business;
• you want a costed plan before committing a budget.

If you already know which task to hand over first, the 14-day sprint is faster.

Two weeks, day by day

Week 1: understand and map

• Day 1: scoping with you. Your typical week, your priorities, the topics that take your time without needing you. We agree on who I meet and which access I need, which you open yourself.
• Days 2 to 4: interviews and observation. I meet your leadership team and the people who run the key flows: sales, invoicing, cash, operations, email, meetings. I look at where your data lives and how it moves, across entities if you have several.
• Day 5: the map. Who does what, how long it takes, with which tools, and what waits for your approval.

Week 2: size and prioritize

• Days 6 to 8: sizing, task by task. What AI can prepare, what must remain a human decision, the effort to build it, the subscription cost.
• Day 9: the prioritized plan. The order of projects, their duration, what each one needs to start.
• Day 10: the debrief, with you and, if you wish, your leadership team.

What you walk away with

1. A map of your week and your teams' weeks. Tasks, time spent, tools and data used, and every point where work stops until you approve.
2. A sizing of what can run without you. For each task: what an agent can take on, what will still need approval, and the build effort. Times are measured where possible and estimated with you otherwise, and every estimate is labeled as one.
3. A prioritized plan. The first projects, the next ones, and what each requires beforehand: connecting data, opening access, naming an owner.
4. The debrief. A session to present, challenge and decide. You keep a document you can execute with me, with your own teams or with another provider.

What I draw on

My proof is internal for now, and I give it with its context. At Oliverlist, founded in 2016 without outside investors, automation tripled revenue per person between 2019 and 2023, before generative AI. Since July 2026, the layer I built for my own work as CEO has handled 200 to 230 runs per business day, 99.4% of them successful. The details are on the agents in production page.

A guaranteed, measured return

The audit ends with a quantified gain target: hours given back, costs avoided or extra revenue, calculated on your data. That target is written into the engagement proposal, and it is guaranteed: if it is not met by the agreed date, I keep working at no charge until it is.

FAQ

How much does the audit cost?

It depends on the size of the company and the number of functions covered. The price is quoted before we start and does not change afterwards. If you sign an engagement within the month, it is deducted.

How much of your time does it take?

From you: the scoping on day one, a few exchanges along the way and the debrief. From your teams: one interview per person involved. The rest of the work is on my side.

Do we need clean data or modern systems?

No. The audit starts from what you have, spreadsheets and inboxes included. My own financial tracking was built in one day from 10 years of spreadsheets, 28,719 transactions in all, and it found 368 duplicates and a VAT discrepancy along the way.

What happens after the debrief?

You decide. The plan is yours: execute it with your teams, run a sprint on the first project, start the 3-month engagement to rebuild a key function, or have me steer it all as your fractional AI director.