Joshua Francis
Case Studies

From Manual Grind to a Repeatable AI Visual Workflow, in Three Weeks

Marketers, designers, and founders were losing hours to manual visual work, outsourcing it, waiting on designers, or fighting AI tools that gave them unusable results. I built a three-week workflow that took them from overwhelmed to producing production-quality visuals themselves, fast. It earned a perfect 100/100 rating, an 89 NPS, and a 74.5% completion rate.

From Manual Grind to a Repeatable AI Visual Workflow, in Three Weeks
100/100
Perfect feedback score from course participants
89 NPS
Net promoter score, a sign of strong advocacy
74.5%
Completion rate, far above the online-course norm

About

Synthminds · on UplimitA three-week AI image-creation course built for Synthminds and delivered on Uplimit, aimed at marketers, designers, and founders who needed strong visuals fast.

Challenge

Non-technical professionals were losing hours to manual visual work or fighting generative AI tools that gave them unusable results, with no repeatable process to follow.

Solution

A focused three-week workflow, refined cohort over cohort, taking participants from zero experience to producing production-quality visuals themselves.

01

The challenge

Everyone I built this for had the same problem. Marketers, designers, and founders needed strong visuals, and creating them was slow, manual, and expensive. They were hiring it out, waiting on designers, or trying to do it themselves and losing hours they didn't have.

Generative AI looked like the way out, but in 2023 it mostly wasn't delivering. Without a process, people got mangled hands, garbled text, and mediocre results, then gave up. The tools were never the problem. They had no repeatable way to use them with intention.

So the gap wasn't a tool. It was a workflow: a structured path from "I've never used this" to "I can produce the visuals my work needs, myself."

They didn't need to become AI experts. They needed a repeatable process, and that's what I built.
02

The approach

I built that workflow as a focused three-week course, designed for the non-technical professional rather than the AI enthusiast. Every lesson had to earn its place by moving someone closer to producing usable visuals on their own.

It built week over week: fundamentals and prompt engineering, then photorealism and design technique, then real professional output, decks, marketing assets, and visual narratives they could use immediately. To serve beginners and power users at once, I ran a dual track: DALL-E as the forgiving core, with optional Midjourney and Leonardo sessions for anyone ready to go further.

And I treated it like a product, not a one-off. After each week I collected feedback and, between cohorts, rewrote what was unclear and added tutorials wherever people got stuck, so the workflow kept getting sharper at the exact points that slowed people down.

The outcome was the relief they came for. Instead of outsourcing visuals or fighting the tools, participants left with a process they could run themselves, fast. It rated a perfect 100/100, earned an 89 NPS, and 74.5% of them finished, the number most online courses never hit.

100/100
Perfect feedback score from course participants
89 NPS
Net promoter score, a sign of strong advocacy
74.5%
Completion rate, far above the online-course norm
Synthminds' expert instruction, guidance, and support were pivotal to me learning a vast world of AI-powered image generation techniques. I learned about color theory, camera lens effects, artistic styles, and so much more that absolutely blew me away. I feel like I truly haven't been leveraging these AI tools nearly enough, so I am incredibly thankful I took this course and got to learn from the industry's best on GenAI.
Brian H. HoughFounder & Software Engineer

Let's get you results like these.

A free 30-minute call. I'll tell you straight what I'd do and whether I can help, no pitch deck.