Selected work
Systems we've actually built.
OriMar Bikes
Problem: Repetitive bike-rental enquiries, availability checks and manual operational administration.
Built: An integrated AI assistant and operations system connecting live bike availability, rental quotes, delivery-zone checks, automated email handling, digital contracts, a central management dashboard and human handoff.
Benefits: Faster customer responses, reduced repetitive administration and 24/7 assistance with bike selection, availability, pricing and delivery. Automated email handling and digital contracts streamline rental operations, while a central dashboard provides visibility and control, allowing staff to focus on customers and bike preparation rather than routine enquiries.
OriMar Living
Problem: Imported property listings were arriving faster than a small team could manually review, optimise and maintain for search engines.
Built: An integrated property management system combining multi-feed filtering and synchronisation, consistent property identification, AI-enhanced descriptions, unique SEO titles and meta descriptions, catalogue-wide duplicate detection, and automated approval, repair and exception workflows.
Benefits: Reduced manual administration, faster property onboarding, consistent listing quality and improved search visibility. A small team can efficiently manage a large property catalogue while retaining editorial control.
Albert's Vinoteca
Problem: Inconsistent product onboarding, incomplete wine catalogue information and limited personalised guidance for customers shopping online.
Built: A structured WooCommerce product catalogue with standardised wine attributes, bottle imagery, tasting notes, food pairings, vegan suitability and SEO metadata. Integrated an AI Sommelier to help customers discover wines through natural conversation, recommend suitable bottles, explain wine characteristics and guide customers towards relevant products.
Benefits: A more consistent and searchable wine catalogue, reduced manual administration, personalised customer recommendations, improved product discovery and a more engaging online shopping experience.
What these prove
Different industries. The same systems mindset.
Live business data
AI and websites become more useful when they can work with the system that actually knows availability, products or properties.
Repeatable operations
The objective is not merely a nicer page. It is less manual checking, less copying and more predictable workflows.
Controlled automation
Useful automation knows where its authority ends. Start supervised, validate the business knowledge and rules, then automate the enquiry types that have earned that trust.