How to start with AI in your company — the first 90 days
A calm, practical plan for the first 90 days: what to look at, what to build first, what to measure, and the four mistakes that kill most first AI projects.
Most companies get stuck at the same point: everyone agrees AI matters, nobody knows what to actually do on Monday morning.
Here is the sequence that works, without a strategy programme and without a six-figure commitment up front.
- Weeks 1–3: understand where the hours go. No tools, no vendors.
- Weeks 4–6: pick one process and define what success means numerically.
- Weeks 7–12: build it, run it in production with a human reviewing.
- Only then decide about the second process.
Days 1–21: find the hours
Before any technology decision, you need an honest picture of where time is spent. Sit with the people doing the work and record each recurring task: how often, how long, which systems, which hand-offs, what goes wrong.
The output is not a slide deck. It is a list of workflows with hours attached — which immediately shows the two or three that are worth attacking.
If you skip this step you will automate whatever is easiest to demo, not whatever is expensive.
Days 22–42: choose one process and define success
Pick the process with high frequency, clear rules and real pain. Then write down the target in numbers: quote turnaround from two days to two hours, invoice processing time down 70%, month-end close from five days to one.
Without that number, the project ends in opinions. With it, the decision about a second stage takes five minutes.
Days 43–90: build it and run it for real
Build the smallest system that runs the whole process end to end, connect it to the systems it needs, and put it into production with a human approving every output for the first weeks.
As the approvals become boring, reduce them to exceptions only. What you have then is not a pilot — it is a working part of the company that already pays back.
The four mistakes that kill first projects
- Starting with technology instead of hours: buying a platform before knowing the workload.
- Pilots with no production data. They always succeed and prove nothing.
- Automating a process the team disagrees about. Fix the disagreement first.
- No owner. A project without one named decision-maker on the client side stalls by week six.
What you need internally
Not an IT department. You need one person who can decide, a few hours per week from the people who know the process, and a willingness to write down how the business actually works — including the parts that are messy.
Everything else — architecture, build, integration, hosting, operation — is the provider's job, and any provider that needs more from you than this is offloading their work onto you.
Frequently asked
- Should we hire an AI specialist first?
- Rarely. In a mid-sized company the first system is best delivered by a team that has done it before; you hire internally later, once there is a system worth maintaining.
- What if we pick the wrong first process?
- Staged delivery limits the damage to one stage. In practice a proper process audit removes most of that risk, because the hours are on paper before anything is chosen.
- How do we get the team on board?
- Involve the people doing the work in the mapping, and be honest about intent: this removes admin, not people. Teams that helped design the system defend it; teams it was done to will quietly work around it.
What this actually means for your company
Imagine the recurring work — quotes, invoices, documents, follow-ups, reporting — simply being done. Not by a bigger team, not by another subscription, but by a system that knows how your business works. Your people stop feeding software and go back to the work you hired them for. That is the whole point.
How this went for me
I ran operations where every day started with the same twenty small tasks. We hired more people, bought more tools, and the admin load still grew faster than the revenue.
The turning point was not a smarter tool. It was writing down how the business actually runs and building one system around it — then letting AI agents work inside that system. Within a few months the daily paperwork was no longer a leadership problem, and we could finally think three months ahead instead of three hours.
That is the only reason DND Systems exists: to do the same thing, properly engineered, for companies that recognise this situation.
Tell us how your business runs today
Write one email describing your day-to-day operation. You get an honest answer on what can be automated, what should stay human, and what a realistic first step would cost. No call centre, no sales sequence.
We reply within one business day, in English, Spanish, French, Portuguese or Russian.
