What it knows.
Your products, customers, numbers, trade terms, brand voice and standards. The information that makes its work specific to your business.
The AI CPG Workshop · Live, four weeks
Put AI to work across your business, so you can move faster, run leaner, and see what matters. Build the systems with me. Leave knowing how to build the next one yourself.
You already know how to run the business
You and your best people know the customers, the product and the standard. Too much of the week goes into getting the work ready.
WHERE YOUR JUDGMENT MATTERS
WHAT FILLS THE SPACE AROUND IT
Growth adds more of that work. Learning to delegate it to AI gives your existing team more capacity.
A different way to grow your capacity
Teach it your business, give it access to the right information, and define what good work looks like. Then turn the process into something you can use again.
“Here’s what we sell, who buys it, and how we talk about it.”
Illustrative course workflowYour products, pricing, customers, trade terms, brand voice and standards, organized into one place AI can read before every workflow that follows.
What’s accurate, what’s sensitive, and what’s worth teaching it first.
“Why do customers buy, and how should we talk about it?”
Illustrative course workflowReasons customers buy and hesitate, counted and quoted from your reviews, support replies and tasting notes. Positioning and draft product-page, sell-sheet and objection-response copy, linked to the customer evidence.
Commercial judgment: what the evidence supports, who cares most, and what is ready to send.
“Which channels changed, and where should I look next?”
Illustrative course workflowA clear view from your selected sources, with checked totals, source dates and gaps made visible.
The context behind the change and the decision about what to do.
“Have the weekly business brief ready before I need it.”
Illustrative course workflowUse a schedule, clear instructions and checks for missing or outdated information.
Direction, review, approval and a response when an exception needs you.
The value goes beyond the first task. You learn how to build this kind of workflow for yourself.
Built in the business I actually run
Just like you, I run a brand and have products to sell, accounts to grow, and a business to operate. AI helps carry the work across it.
scheduled AI jobs
behind the operation
I hold the judgment calls. Anything that spends money or leaves the company still stops at me. I won’t put NØRSE CØDE’s real numbers on a public page. That’s a line every operator will recognize.

The method behind the builds
Every workflow starts with the same three parts. Learn how they fit together, then apply them wherever your business needs more capacity.
Your products, customers, numbers, trade terms, brand voice and standards. The information that makes its work specific to your business.
The files, reports, resources and knowledge, and the connected systems it needs to do the job. Start with what you have and connect what the workflow actually needs.
Your repeatable instructions, examples and checks. A process you can inspect, improve and use again with new inputs.
Four weeks of building together
We build your company second brain first. Then we put it to work: turning customer feedback into sales copy, your numbers into a dashboard, and a workflow onto a schedule that runs without you starting it.
Learn how to turn your products, customers, terms and standards into a company second brain, so AI has what it needs to do useful work you can check against the facts.
Your next step: Use the same foundation whenever you build another workflow.
A shared starting point
for the work that follows.
Learn how to turn what your customers already said into the words you sell with. Use your company second brain, reviews, support replies and tasting notes to uncover why customers buy and why they hesitate, then turn that evidence into positioning and draft copy for your product page, sell sheet and recurring objections.
Your next step: Turn another task you know how to do into work AI can prepare to your standards.
Learn how to bring a focused set of business reports into one dashboard. Connect the information, check the totals, and see what changed, so you spend your time deciding what to do with the numbers.
Your next step: Add sources and useful views as your business needs them.
Checked totals.
Clear sources. Visible gaps.
Learn how to take a workflow you have checked and put it on a schedule. Define the review points, handle missing inputs, and know how to inspect, pause and improve the job.
Your next step: Use the same approach to keep other recurring work moving.
Built to help you implement
You get the tools and the knowledge to put AI to work in your business, and the help to make it stick.
See the build, understand the decisions behind it, and start yours. Every teaching session is recorded so you can revisit the work.
Bring your questions, your output and the thing that didn’t work. We address common blockers and help you take the next step.
Use the templates and worked examples on your own inputs. Practice between calls, then return with something concrete to improve.
The build after the course
You’ll know how to point AI at the next job the same way you did these four: gather the context, define what good looks like, and check the result against the facts. A supplier scorecard. A launch-readiness checklist. A weekly report that used to eat your Friday afternoon.

Smári Ásmundsson Founder · Operator · Your instructor
A note from Smári
For more than 15 years, I was an advertising photographer, creating work for brands including BMW, Volkswagen, Rolex, Land Rover, BASF, Four Seasons, Hartford Insurance and the Pacific Symphony Orchestra. The job was to understand the brief, see the possibility, and hold the work to a standard.
In 2011, I founded Smári Organics. I took Icelandic skyr from my home kitchen into national retail. I learned what it takes to get a physical product made, sold and onto shelves.
Today, I run my second brand, NØRSE CØDE, and build AI systems that help us operate. Through Nordic Navigators, I teach other operators how to put that capability into their own businesses.
You already have the expertise. I’ll help you turn it into work AI can do.
BMW · Volkswagen · Rolex · Land Rover · BASF · Four Seasons · Hartford Insurance · Pacific Symphony Orchestra
Smári Organics · NØRSE CØDE
Who this is for
You run a seven to eight-figure CPG brand. You know where work gets stuck. You want the ability to do something about it yourself, and you want to implement the systems that help you grow faster.
You want practical AI implementation in a real business.
You’ll make time to apply the work between sessions.
October 12 to November 2, 2026
Build useful systems for your business, and the ability to keep building after the course.
Secure checkout via Stripe. A paid Claude plan is required; the $100/month Max plan is recommended.
Move to a later run any time before the first session, for any reason.
Before you join
What to expect, what to bring,
and how the course works.
No coding experience is required. You’ll learn to give AI the context, instructions and standards it needs, then review what it produces.
The course moves from individual conversations to reusable workflows built around your business. You’ll organize the context, work with your own inputs, define quality checks and put a working process on a schedule.
Each week includes one hour of live teaching and a separate one-hour office hour. Plan for about 2 hours between sessions to gather inputs, apply the builds and test them on your own business.
A paid Claude plan (the $100/month Max plan is recommended) and a computer you can use for the builds. You’ll receive a preparation checklist before the first session. Use materials you are authorized to share with your AI tools. Paid software and any optional hosting or API costs are separate.
That’s expected, every business runs different systems. You learn to identify sources, map definitions and check the result. We start the dashboard with two sources and two comparable periods. Exports are a valid starting point; complex integrations and full ERP or demand-planning implementations are outside the four-session scope.
You define the boundary. The course begins with reading, preparing and producing work for review. You’ll learn where approval belongs and what to do when information is missing. Sending external messages or making commercial commitments is not required to complete the builds.
Teaching sessions are recorded, and the templates and worked examples are yours to keep. Catch up before the next lesson and bring implementation questions to that week’s office hour. The sessions build on each other, so applying the work between calls matters.
The shared teaching build uses a demonstration brand. You apply it to your own information in your own account. If you bring a file for help during office hours, you choose what to share and can remove sensitive details first.
You can move to a later run any time before the first session, for any reason.
Your expertise is the starting point
Four weeks. Useful systems. A method you can keep applying.
Build it with me in October