I turn scattered, unofficial AI use into clear, safe workflows your whole firm can trust — designed with the people who'll actually run them.
30 minutes. No pitch. You'll leave with the two places I'd look first.
Microsoft & LinkedIn Work Trend Index · McKinsey State of AI
None of this is a technology problem. It's an operations problem — which is good news, because operations can be fixed.
Saving hours is the easy part. The return comes from what your firm does with them.
Some of that time goes back into billable work. The rest is where firms grow: leaders with room to think and strategise, teams learning instead of formatting, better client conversations. The roadmap plans for both — the hours saved, and the hours reinvested.
Sixty minutes on goals and worries — in your words, not mine.
Leadership to front line: where time goes, what they'd never trust a machine with.
Priorities, owners, costs, measures. Yours to keep, with or without me.
Built with the people who'll use it — the only version that survives a busy Thursday.
Prices are on the page because you shouldn't need a sales call to know if this fits your budget.
Know exactly where AI fits your firm — and what to do first.
One or two workflows built — and actually used.
A standing seat in your firm. We agree the scope first — the price follows from it, not the other way round.
Sprint pricing fixed before work starts · software costs never marked up
Twelve possible AI projects, shortlisted to four. The partners wanted automation; the team wanted their afternoons back. Same problem, it turned out.
The same client information typed three times, and drafting meant an afternoon of copy-paste. We're rebuilding it from intake onwards.
No "saved 400 hours!" claims here. Baselines are being measured now; the before/after numbers get published once they're real. If a consultant quotes results they can't show their working for — ask for the baseline.
My background is project management and digital systems: a postgraduate degree in project management, years building websites and operational tools for businesses, and leadership roles in education — including running the academic programme at an international summer camp each year. Today that combination goes into AI operations work for firms in Singapore, Australia and the US.
The through-line: technology fails in firms for human reasons, not technical ones. Delivery discipline gets a system built; training and change work get it used. I do both — and I build alongside my business partner Raghu, who shares the deeper technical work with me. Grew up in Botswana, Romanian parents, educated across five continents — which is why working across cultures and time zones is normal here, not a stretch.
Best fit: a service firm of 10–100 people doing $1M–$10M — professional services, architecture, legal, education, media, agencies — with someone internally who can own what we improve. Not a fit: you want a chatbot by Friday, or AI making decisions with no human in the path. And if the budget isn't there this year, say so — I'll point you to free resources instead.
No. You're not behind — only about one in five organisations has actually redesigned how work gets done with AI. We start from where your firm actually is.
It's the first thing we design for. The audit draws the line in writing: what AI may touch, what it never touches, where a human signs off.
The tools were probably never the problem — the rollout was. When the people doing the work help design the workflow, it stops being something done to them.
Yes — it's designed for exactly that. Some firms run it themselves and come back later for the build. That's a fine outcome.
The AI Practical Brief: one practical idea per issue, fortnightly, for people who are done with AI hype. 17 issues in.
Thirty minutes, free. You'll leave with the two places I'd look first — whether or not we ever work together.
Book your free practice reviewPrefer email? hello@lilianaoperations.com — I reply within a day