Fewer manual hours, honestly counted
Practical AI woven into real business workflows — prompt libraries, templates, SOPs and training — run daily across my own six ventures before ever being recommended to yours.
per week, at peak
— down from 3–7 days
took 3–5 hours
workflows daily
Work moving through, unattended
Automation is not a chatbot on the homepage — it is the repetition quietly stopping. Drag it about.
Where the manual hours actually go
Four patterns show up in almost every review. If one of them made you wince, that's your starting number.
A human is the template
The same letter, proposal or report gets recreated from scratch every time it's needed — by somebody senior, at a salary, from memory.
Friday is for reporting
Someone assembles the same numbers into the same document every week. A report that a framework could produce in minutes takes the afternoon.
The same email, forever
Most of the inbox is the same dozen questions, each answered from scratch, politely, by someone with better things to do. A tested reply library answers them in your voice in seconds.
The chatbot nobody uses
You've bought AI once already. It sits on the homepage answering questions no one asked, while the genuinely repetitive work carries on by hand behind it.
Automation that removes work, not automation that looks impressive
There's a great deal of AI theatre about. Meanwhile, in a typical service business, ten or more hours a week disappear into the repetitive middle — recreating the same documents, chasing the same information, assembling the same Friday report. None of that needs a chatbot on the homepage. It needs the repetition to stop being a job a human does from scratch every time.
My version is unglamorous, and it works: prompt libraries, document templates, standard operating procedures and AI-assisted drafting workflows, built around how the business already operates and tested inside it before handover. Routine output becomes a matter of updating a framework rather than starting from blank — with a human reviewing everything that matters. And every engagement is documented, so you're not renting my memory.
I'm pragmatic rather than tool-driven. Where AI genuinely earns its place — drafting, research, analysis, iteration — I use it daily across my own companies. Where a spreadsheet, template or checklist solves the problem better, that's what you'll get. My clients aren't buying AI — they're buying time, better systems and improved profitability.
Measured on my own operations before yours
Every workflow I recommend has already earned its keep inside my own companies. I don't automate to replace people — I automate repetitive work so time goes where it creates value: solving problems, building relationships and making better decisions. The numbers below are floors, not marketing.
10–15 hours a week, returned
Compliance tracking, contractor coordination, resident communication and legal correspondence across a 77-apartment portfolio — rebuilt as templates, checklists and centralised records. Most routine tasks became updating a framework rather than starting again, and missed actions largely disappeared.
From a week to a day
Repeat product enquiries used to mean re-finding suppliers, fresh quotations and renegotiation — typically 3–7 days. A documented supplier network and repeatable procurement workflow brought most repeat requests down to same-day to 24 hours, with trusted manufacturers already vetted.
An afternoon, down to 90 minutes
A hotel review, restaurant write-up or music release once took 3–5 hours of drafting, editing and asset production. Structured workflows, reusable frameworks and AI-assisted drafting deliver comparable quality in 45–90 minutes — with final human editing and approval, every time.
Only what I actually run
The tools on this page are the ones I use daily and would happily be quizzed on: ChatGPT, Claude and OpenAI image tooling for drafting, research and creative work; Kimi for Chinese-language client work — a two-cities detail most UK consultants can't offer; Adobe Firefly and Kling AI on the generative side; the Adobe Creative Cloud apps, Photoshop among them; Suno in the music workflow; CapCut Pro and Canva for content; Excel, Word and Google Workspace doing the unglamorous heavy lifting; the Instagram, TikTok and YouTube creator stacks; DistroKid and Spotify for Artists on the release side; CRM, supplier-database and quotation workflows underneath it all.
There's a longer list of platforms I know but don't lean on daily — and on principle it stays off this page. I only recommend what I run. If your problem genuinely needs heavy integration engineering, I'll say so plainly and help you brief the right specialist rather than sell you my learning curve.
What an engagement hands over
Three artefacts, every time. The system ends up owned by the business, not rented from me.
The review report
How the business actually operates, where the repetitive hours go, quick wins, and a prioritised roadmap — useful even if nothing further gets built.
The working system
Prompt libraries, templates, SOPs and workflows — built for your business, tested inside it, refined on real feedback before handover.
Trained people, full documentation
Owner and key staff shown how to run it, everything written down, and a system designed to keep delivering long after I've left.
Start with the review
Nothing gets built until we understand what it's replacing and what that's currently costing you.
Review
Time inside the business, documenting how work actually flows and where the repetitive hours go. Deliverables: workflow assessment, opportunity report, quick wins, roadmap.
Build
Prompt libraries, templates, SOPs and workflows tailored to the business — never technology forced into a process where it adds nothing.
Implement
Integrated into the day-to-day operation, tested against real work, and refined on feedback until it's genuinely easier than the old way.
Train & hand over
Your people trained, documentation delivered, system owned by the business. The goal is independence, not dependence.
Asked on almost every call
If your question isn't here, it's probably the first one on the call.
What does it cost, and what does it save?
The review is a fixed fee. Every recommendation is placed next to the manual time it removes, so the payback is in front of you before you commit to anything. If a process isn't worth touching, the report says so and you leave it alone.
Which tools do you recommend?
The ones I run daily myself — ChatGPT and OpenAI tooling, structured templates, spreadsheets, and the creator stack that publishes across my own brands. No commissions, no platform to push. If your problem needs heavy integration engineering, I'll say so and help you brief the right specialist.
Will my team actually use it?
That's the design constraint everything else answers to. The workflows are built around how your business already operates, tested inside it, and refined until they're genuinely easier than the old way — because systems people actually use beat systems that impress consultants.
Is our data safe?
The work happens in your accounts and your files, and nothing sensitive goes into any tool outside your policies. What's handled, where, and by what is agreed in writing as part of the scope — not offered as a verbal reassurance.
Do we actually need AI?
Quite possibly not everywhere. A respectable share of what's sold as AI is a template with better marketing — and when a spreadsheet, checklist or document framework is the right answer, it's cheaper, faster and breaks less. I'll tell you which yours is before anything gets built.
What are you still doing by hand?
Describe the task that eats the most hours — roughly how it works and roughly how often. I'll tell you honestly whether it's worth systemising, and what I'd expect it to give you back.
hello@pikotw.com