Loxord is a luxury fashion house in Cairo. It makes leather objects meant to be kept: the Nougat Tote, cut from vegetable-tanned leather in five colours, made in Cairo to preorder at LE 18,000 and shipped worldwide. The house describes its work as building "companions for future histories". It describes its service in nine words.
"The house answers. Calmly, and in order of arrival."
Loxord, contact page
At that price, with a wait between payment and delivery, service is not a department. It is part of what the customer bought.
Service is part of the object
A customer who has paid for a tote that is still being made has questions a tracking link cannot answer. When will mine be ready? Will the leather darken? What if it arrives and it is not for me? Can it be repaired in ten years?
Loxord has written an exact answer to each. Leather that darkens where hands return is "evidence of life", not damage. A customer has fourteen days from delivery to say an object is not for them. A defect in making or materials is covered for two years. Repair is "part of ownership": the customer writes with the object's archive code and a photograph, and the object returns home for its work.
For a house like this, white-glove service comes down to four things:
- The answer is the house's answer. Not a general one about leather, and never a promise the policies do not make.
- It sounds like the house. Calm, exact, unhurried.
- The customer is never asked twice. The order number, the archive code and the photograph are collected once and travel with the conversation.
- A person arrives when a person is needed. Repairs, faults and anything that takes judgement belong to the team.
A chatbot measured on how many conversations it keeps away from people fails all four. A luxury customer does not want to be deflected. They want to be answered.
An agent tailored to the house's workflow
Loxord added Aide to its Shopify store in September 2026. Twenty-six days in, the house wrote this on the Shopify App Store:
"We were able to tailor this agent to fit our store's exact workflow. The AI responses are sharp and accurate, and the fallback feature of collecting emails when it gets stuck makes it seamless for our human team to step in and close the loop."
Loxord, Shopify App Store review, October 2026
Tailor is the word that matters. In Aide, a team shapes the agent in plain language, the way it would brief a new colleague:
- Knowledge is the store's own: its policies, its pages and its Shopify products.
- Instructions are rules that hold in every conversation, such as never promising a delivery date.
- Procedures say what to do in one situation: what to ask before a repair request reaches the team, when to hand off, how the conversation should end.
- Tone and identity hold the voice.
Once a customer shows an order is theirs, by signing in to their store account or giving the order number and the email on it, Aide answers about that order: what was bought, whether it has shipped, where it is. The check comes first, so nobody hears about an order that is not theirs.
When a person should close the loop
The second half of Loxord's review is about the moment the AI stops. Aide is built to stop well. Every reply is checked before it is sent. A wrong price or order detail, a promise the policies do not back, or a reply that does not answer the question is rewritten once. If that does not fix it, Aide asks a clarifying question or hands the conversation to the team.
When it cannot answer, it says so in the customer's own terms and offers the next step. A customer who asks for a person gets one, without the AI answering first.
The handoff keeps what the customer already gave. A procedure can ask for the order number, the details and a photograph, one message at a time, before it hands off, so the team opens the conversation with everything in it. Aide reads the photographs a customer sends and says only what a photograph shows. In Aide Chat, it collects the customer's email, the team's reply reaches them after they have left the page, and their answer lands in the same conversation.
And when a customer asks whether they are talking to a person, the answer is fixed and true. No setting or instruction can make Aide claim to be human. A house that states where its objects are made as fact can state that as fact too.
Quality that is checked, not assumed
The usual measure for AI in support is deflection: how many conversations never reached a person. It rewards the wrong thing. Aide's measure is whether a reply was acceptable.
Each agent has a Quality page with a pass or fail headline against a written quality bar. Every night a sample of the agent's replies and of the team's own replies is graded on one scorecard, so the page shows the agent beside the team's median and its best people on the same topics. A house can see whether the agent writes as well as its own team before it lets it reply. A teammate can mark a reviewed draft Good or Wrong, and every Wrong draft becomes a test the agent is replayed against each night. This is the working part of Aide's Agent Governance Engine: the agent acts inside rules the team wrote, and its work is graded where the team can see it.
The same loop leaves the team knowing more, not less. Each night Aide lists what customers asked that the house's knowledge did not cover, with the conversations behind each gap and the text it proposes. Where teammates answered the same question two ways, it asks which answer is policy. Nothing a person wrote changes without a person. Over a season, the house's written knowledge of its own customers grows: what they ask before a preorder, where a policy reads two ways, which answers needed a human.
What a house can ask of its AI
Loxord's objects are built to change well and to be repaired rather than replaced. The service around them is held to the same idea: exact answers in the house's voice, a person when a person is needed, and a record that gets better with use.
Luxury service is not more conversations kept away from people. It is every customer answered the way the house would answer.
Loxord's words and policies are quoted from loxord.com as published in November 2026. The review is from the Shopify App Store, dated October 18, 2026. The sections on how Aide works describe the product, not Loxord's own settings.