Launching an AI-powered engineering platform from beta to market-ready product
About the project
Vixiv is a Cincinnati-based AI-driven design platform created to revolutionize how engineers optimize parts for additive manufacturing. Instead of months of manual trial and error, engineers upload a CAD model, set constraints like load and material, and receive dozens of validated design variations in minutes.
The platform produces functionally graded lattice structures trained on real-world destructive testing data — not synthetic simulations. Aiming at early- to mid-career engineers in aerospace, automotive, and medical fields, Vixiv connects the gap between initial concept and production-ready geometry with a speed that traditional tools cannot provide.
Challenge
Vixiv had a functional beta, no public-facing presence, and a hard conference deadline weeks away. The team needed a polished promotional video, a landing page to capture interested engineers, and then a complete product interface redesign to support external beta users — all within a four-month engagement.
Our process
We moved from a high-pressure conference sprint into sustained product design and development, shaping the public brand and the core engineering tool.
{{dropdowns-1}}
{{slider-1}}
Results
Vixiv launched its public beta on schedule, with the landing page and promo video ready for the conference deadline. The redesigned product interface gave external beta users a clear, self-guided path from model upload through AI-generated optimization — replacing what had been an internal-only alpha. The platform gained press coverage across industry publications, including Fabbaloo and Additive Manufacturing Media, strengthening Vixiv’s visibility in generative design.
Our approach
We viewed an urgent conference deadline as a forcing function, delivering public assets first and then applying that momentum to deeper product design and engineering work.
{{approach}}
01 · Discovery & research
We mapped Vixiv’s optimization workflow against 5 competitors, including nTop, Autodesk, and Altair Sulis. This analysis clarified where an AI-first approach created distinct UX opportunities that manual-heavy rivals could not replicate.
Deliverables
- Competitor benchmarking
- Workflow audit
- Conference timeline plan
02 · UX/UI design
We created the full product interface — from login and registration through the 3D workspace where engineers set constraints and review AI-generated results. Iterative feedback led to key decisions, such as maximizing 3D viewing space.
Deliverables
- 3D workspace screens
- Account & billing pages
- History & file management
03 · Web design & development
We designed and built a 3-screen Webflow landing page with a splash screen to convert engineers post-conference. A motion designer created a 2-min animated video using demo screens showing the beta's full capabilities.
Deliverables
- Landing page design
- Webflow development
- Promotional video
04 · Front-end development & QA
The front-end team built the product using TypeScript, React, TailwindCSS, and Three.js, with a 3D specialist on the model viewer. Our QA specialist tested browser compatibility, console errors, image optimization, and 404 errors for stability.
Deliverables
- React product build
- Three.js 3D integration
- Pre-release QA validation
Team
A cross-functional team worked in close sync, continuously adapting to new insights from classroom trials and robot behavior.
- 1 Project Manager
- 2 Brand Designers
- 2 UI/UX Designers
- 2 Web Designers
- 1 Front-end Developer
- 1 Back-end Developer
- 1 QA Engineer
Technologies
Our stack supported offline interaction, hardware communication, scalable content management, and smooth cross-platform UI.
Team
A cross-functional team of designers, developers, and a 3D specialist worked in parallel across the product interface and public-facing assets.
- 1 Project Manager
- 1 Lead Designer
- 1 Product Designer
- 1 Web Designer
- 2 Frontend Developers
- 1 3D Developer
- 1 QA Engineer
Technologies
We paired Webflow for rapid landing page delivery with a React and Three.js stack for the interactive 3D product, keeping both tracks moving in parallel.
