Hey {{first_name | Friend}} 👋

This week, we’re shining our light on something special happening inside our community.

Lighthouse alum, Dario van Houwelingen, has been fervently building a full AI visualization system: a bridge between creative experimentation and commercial production.

What started as a test inside ComfyUI has morphed into a functional tool. 

One that lets marketing and design teams generate campaign-ready product concepts from 3D assets, complete with lighting, composition, and photographic realism.

Here’s how he built it and what every creative technologist and AI connoisseur can learn from the process.

Dario began by crafting a complete 3D dataset himself.

He modeled and rendered a high-end motorcycle from dozens of camera angles, environments, and lighting setups; more than 20 backplates and HDRIs in total.

Each image served as a foundation for consistency: form, shadow, and surface behavior.

This dataset acts as a visual fingerprint, going beyond simple training material: it captures how real-world photography behaves under variation.

It’s extensive work to train a model like this, and while that level of effort isn’t always necessary, working at the highest quality of product visualization demands the highest quality assets.

Details truly matter, and we get into that next.

After captioning and refining his data, Dario trained multiple models before landing on a LyCORIS Flux Dev configuration, selected for its ability to retain fine surface detail while allowing stylistic flexibility.

This balance (control and nuance) became the technical backbone of the workflow.

Once the model was flexible and stable, Dario built a ComfyUI pipeline that merges deep control with intuitive guidance.

He rendered custom ControlNet images from his 3D asset to lock camera perspectives and added a Claude LLM node that expands short prompts into studio-grade compositions.

Example:

User input → “motorcycle on a mountain road”

Claude expands it to → “[TRIGGERWORD], a sleek motorcycle resting on a winding mountain road at golden hour, chrome reflecting warm light, cinematic composition, 35mm lens, high contrast between cool shadows and sunlit peaks.”

This step turned technical complexity into creative fluidity. Fortunately for marketing and product teams, anyone can use it.

With help from Flowscale and Bolt, Dario deployed the pipeline through a cloud-based UI, complete with camera icons, preset toggles, and real-time cloud GPU execution.

This resulted in a functional prototype that marketing teams can use directly.

And this turns a workflow into a fully-fledged app.

Now in testing with marketing departments and product studios, Dario’s system is being positioned as a concepting tool for automotive, cosmetics, and consumer brands.

Yes, it’s designed for speed, but it’s also built for substance:

If this workflow resonates with you or you have any further questions, please reach out.

It’s as easy as scheduling a call.

We would love to get your feedback on what works and what doesn't regarding this process.

Keen to follow Dario’s lead, without building a full workflow?

Try this simple exercise inspired by his use of Claude for prompt enhancement.

Do you see how Claude translates vague ideas into production-ready language just as Dario’s workflow does automatically?

This simple loop (dataset → control → iteration) is how workflows turn into products.

Feel free to share your outcome with us by email, or tag us on LinkedIn if this helps!

Applications are officially open for all three Lighthouse AI Academy tracks:

From zero to clean, readable visual workflows — no prior node experience required.

Who it’s for: Artists, designers, and creatives building their first AI foundations.

👉 Get Started on Your Application:

From prototype to pipeline — batching, automation, quality checks, repeatability.

Who it’s for: Technical artists, pipeline builders, VFX supervisors, and experienced ComfyUI users.

👉 Get Started on Your Application:

Design systems (not just outputs): workflow design, team integration, ethics, and portfolio projects.

Who it’s for: Creative directors, studio leads, strategists, and innovation teams bringing AI into their practice.

👉 Get Started on Your Application:

If someone in your network enrolls, they get 10% off, and you receive 10% back on any Lighthouse course.

It’s that simple and awesome. 

One alum recently referred five students and earned €1,200 in tuition credits: great proof that sharing is caring really does pay off.

Write your referee’s name when applying, and we’ll handle the rest.

Community-driven growth for spreading colorful creativity.

Spread the word: Early Bird pricing is live for any 2026 cohort. Lock your seat early and SAVE 10% on tuition while you can! 😉

Projects like this are stark proof that creative technology can turn curiosity into something real… and human.

Dario’s build is more than a workflow. It’s a reminder that when you treat your process like part of the craft (not just what happens behind it), new possibilities abound.

We hope you found the breakdown of this experiment becoming a product — not through luck, but through clarity, iteration, and care — as enlightening as we did.

That’s the next phase of creativity: tools that think with us, not for us.

If you can explain your process, you can share it — and if you can share it, you can build something that lasts.

That’s it for now: Thanks for reading this edition of Get Comfy with LAIA and building this new era with us.

If you’re ready to push your next experiment further or simply begin one afresh, applications for our 2026 cohorts are open, and don’t forget to invite a friend through the Affiliate Program!

Step by step, node by node — and away we go.

Keep creating and always remember to have fun.

— Adam & the Lighthouse AI Academy Team ☀️