AI for your organisation
Most processes were built for people, who fill the gaps from memory. AI can’t. I rebuild how the work flows so AI can carry the repetitive parts, then train your team to run it. Same team. Far more done.
Computer-science engineer MBA, IIM Ahmedabad Founder of LEAD
People I’ve trained and coached work at
Why AI disappoints
The monthly report
“AI just doesn’t work for us.”
Eight departments send their monthly figures, each in its own format: a spreadsheet, a PDF, an email with a table pasted in. One team spends days stitching them together. Someone hands the pile to AI. It makes a mess of it.
The two exports
“It doesn’t even get the basics.”
Two systems, two exports. One column says Cod. Cliente, the other says Account ID. Laura in finance has always known they’re the same customer. Nobody ever told the AI. It can’t match a single row.
What the AI sees
Cod. Cliente Account ID
Same customer. Two systems. One rule nobody wrote down.
Processes grow around people, and people carry the rules in their heads: which file is current, which name means what, who needs it next. AI can’t read minds.
So I start with the process. Write that knowledge into how the work flows, and the same tools start delivering.
What 30% is worth
That’s what I see again and again in my training programmes, often within a couple of weeks. Not from longer hours. From AI taking over the repetitive work, so people spend their time where their expertise counts.
Two ways to work together
Not sure which you need? Most people aren’t. Tell me how the work runs today and I’ll suggest the smaller step first.
01 Team training
I start by learning how your team works. Then I make a few small changes so AI fits the process, and train your people on their real tasks: the reports, emails, analyses and documents they handle every week.
No generic prompt course. Results show up in weeks, not quarters.
02 Process redesign
When the work depends on what’s in people’s heads, training alone won’t fix it. We take one shared process, map how it really runs and rebuild it so AI carries the repetitive parts.
I do the architect’s work: which steps stay with people, which AI assists, which run on their own. Then I teach each person the part they’ll run.
Bigger, multi-team redesigns are scoped with you.
Talk about a process
In practice
In one organisation, a new department was being formed: a new org chart, new meetings and minutes, new processes, new people to onboard, training to arrange.
We started with the AI tools they already had and how they were connected. I taught the team those tools. Then we mapped the work, put it into one clear structure and automated the manual, repetitive steps.
It made the job of building that department immensely easier.
The architect’s drawing New department · coordination
01 Trigger
02 Capture
03 One structure
04 Put to work
05 Decide
Then each person learns the level their part needs: Operator for the daily AI-assisted steps, Builder for the automations, Starter for newcomers. The Architect part is mine.
Your tools, your rules
In my own business I use every kind of AI. A bank, an insurer or a pharma company can’t, and shouldn’t. So I work inside your security rules, with the tools you already have.
Copilot is a good example. Give it a clear task, the right context and a tight scope, and it delivers. Knowing exactly what it can do is half the job.
…I used the ideas we discussed to implement end-to-end AI-powered processes that are giving clear returns.
Bring one. In a free 30-minute call we’ll look at how it runs today, and whether training or a redesign makes more sense.
Talk about your team
