AI workflows · Courses · Growth
I'm Nabil Asofi. I build AI workflows and automations, teach the systems behind them, coach teams on products they've already built, and run the SEO and growth programmes that feed them.
Proof of concept
The Rank-and-Cite Monitor from the library, executing unattended. Not a slide about automation — the automation.
This workflow runs in 6 steps:
01 — What I do
Buy the artifact, learn the skill, or hand me the programme. They stack.
Automations that replace a repeating job you or your team currently do by hand. You get the workflow, the config, and a walkthrough — not a strategy deck.
Browse workflows →The same systems, taught so your team can build and maintain them without me. Foundation through advanced, built from client work rather than theory.
See the courses →Programmes run for you: technical SEO, content that gets cited, paid and organic acquisition, and a lead pipeline that is measured rather than admired.
See services →02 — The library
Each one replaces a specific repeating job. Setup time is the honest number, measured against real data.
Turn a target query into a sectioned, sourced writing brief — before anyone opens a document.
Outcome A brief with the sub-questions the SERP implies, the gaps nobody covered, and a word budget per section.
Check every week whether AI answers name you — and get told the moment that changes.
Outcome A weekly diff: which queries you are cited on, which you lost, and who took the slot.
Score every inbound lead on fit, route it to the right person, and draft the reply before anyone reads it.
Outcome Every enquiry tagged, scored, routed, and answered with a draft a human only has to approve.
03 — Method
Four steps. No discovery phase that exists to justify a retainer.
We find the task that happens weekly, costs hours, and follows the same shape every time. That is the only thing worth automating.
Week 1No sandbox demos. The workflow runs on your actual content, your actual inbox, your actual product feed — because that is where it breaks.
Week 1–2You get the workflow, the config, the failure modes, and a recording of the build. If you can read a spreadsheet, you can operate it.
Week 2Monitors catch the quiet failures — the feed that stopped syncing, the rankings that slipped, the leads nobody answered. You hear about it before your customers do.
Ongoing04 — Courses
Built from client work. If a lesson hasn't survived contact with a real project, it isn't in the course.
Build the workflows in the library yourself — and know why each instruction is in the prompt.
Pricing to be announced
Details →Move your search programme from reading dashboards to responding to changes.
Pricing to be announced
Details →Design the acquisition, conversion and retention system as one thing, because that is how it breaks.
Pricing to be announced
Details →05 — Journal
AI news that affects how you work, and the search shifts worth acting on. Written to be read, structured to be cited.
06 — Offers
40% off the first cohort of Deployable AI Workflows. Twenty-five seats, then the price is the price.
NASOFI40 No. Every workflow ships with a written setup and a recording of it being built against real data. If you can follow a recipe and read a table, you can run it. Where a workflow needs a developer, I say so on the workflow page before you download it.
The workflow file itself, the configuration and any prompts it depends on, a setup doc with the exact steps, and a short note on the failure modes to watch for. No upsell gate inside the file.
That is most of the consulting work. The library covers the common jobs; the interesting ones are specific to your operation. Start a project and describe the task that repeats.
As a programme with a named owner and a monthly review, not a one-off audit that sits in a folder. Technical fixes, content that earns citations, acquisition and conversion are treated as one system because that is how they fail — together.