Orium

Optimized Storefront Builder

Orium, a leading composable commerce consultancy, wanted to build an AI powered storefront builder that automated copywriting, design, and A/B testing for retail brands. Ollon led development of the platform, called Path, from UX design and prototyping to production release. Path is now being piloted by two large retailers, adjusting layouts automatically to optimize each brand's experience.

Situation

Orium is North America's leading composable commerce consultancy and systems integrator, supporting how retail brands connect with their customers. They had an idea to create an innovative new storefront builder platform with marketing functionality powered through AI.

Solution

Ollon supported Orium along the development journey, from UX design, to prototyping and building flexible storefront generator. During beta testing, Ollon worked with the Orium team to continue to enhance the product, Path.

Results

With the objective of automating "people activity" such as copywriting, design and A/B testing to optimize landing pages, Path tracks usage and quickly adjusts and changes based on an optimized experience for brand customers, ultimately saving clients' time and money. The platform is currently being piloted by two large retailers.

Technologies

NextJS, Vercel, MongoDB

Our Expertise in Action

Custom application development

Orium wanted a functioning storefront builder that could take a brand's information and generate a working online store. Ollon built the application end to end, architecting the NextJS front end, the Vercel deployment pipeline, and the MongoDB data layer that stored each generated storefront's structure and content. That full stack build is what let the generation logic actually run as a deployed, working product rather than a design concept. The result was a functioning application covering both the visual storefront and the infrastructure running underneath it.

AI-powered content generation

Orium's storefront builder aimed to generate more than layout, brand aligned copy and design decisions produced automatically from a user's input, at a time when generative AI content tools were still new and largely unproven in commerce. The challenge was keeping that generated copy and those design choices specific to each brand rather than reading as generic AI output, using only a brand's product, values, and category as the signal to steer the generation. Ollon applied early generative AI techniques to both layout and content creation to meet that bar. The result let a brand describe itself and receive a generated starting point that reflected its own identity, not a generic template.

MVP development and product-market fit

Orium needed to test whether an AI generated storefront builder had a real market before committing to a full build. Ollon built the MVP version of the builder used for these evaluations, including the template system and the input driven flow that generated a starting storefront from a brand's information. The MVP gave Orium a working product to present directly in investor meetings, where the team could gather firsthand reactions to how the concept might fit into a brand's existing marketing process.