DTI helps brand apps in ChatGPT, Claude and Gemini lead expert conversations, understand how customers choose, and use a portable consumer model.
Brands gain insight into why their products are chosen or rejected.
Explore DTI integration optionsAI can hold a conversation, explain features, compare products and make a persuasive case for a recommendation. But that is not enough to guide someone consistently toward a confident decision.
The ability to reason about a choice does not mean an app structures its consultation like a professional expert.
An expert uncovers what matters, keeps the reasons for a choice in focus, and helps resolve conflicting priorities. Their recommendation reflects a coherent understanding of the customer and their task – not just the latest message.
A customer profile containing saved attributes, preferences and past outcomes does not, by itself, provide this way of working.
DTI is being built to give AI apps the depth of a highly skilled personal consultant in the luxury segment.
One watch for everyday wear. A budget of around €5,000.
A conventional selection based on size, style and brands produced no options that sparked visual interest.
Status through design and craftsmanship, without a loud logo. Recognition among connoisseurs, rather than complete obscurity.
Better to wait than settle for a weak compromise.
“Stop searching by silhouette. Look for the material and craftsmanship of the dial.”
“The absence of a strong ‘no’ is not yet a ‘yes’.”
The profile clarified the apparent contradiction: status mattered, but overt branding did not.
The outcome: three alternatives the customer liked, a step-up option, and a recommendation to revisit a variation of the reference model.
Based on a real watch purchase. Context was gathered using the DTI methodology and applied by a specialised AI agent without retraining.
More than a list of suitable options: support in reaching a confident decision.
Customers understand why an offer fits their task, what they gain and which trade-offs they accept. The recommendation makes the desired outcome tangible, in the language of their own values.
With the customer’s permission, their preferences and experience can inform future conversations with connected apps.
DTI helps the brand’s app communicate the value of an offer to each customer, while giving the brand analytics on why products are chosen or rejected, which alternatives win, and which needs remain unmet.
These insights support assortment evaluation, reveal competition between a brand’s own products, and inform product, positioning and offer decisions.
DTI adds an expert method and Consumer Intelligence to the brand’s app.
The developer retains their product, catalogue and commercial workflows.
Swipe horizontally to view the full diagram.
One expert method across brand apps. Individual consumer models that customers can carry between connected apps. No separate product version for each client.
Both routes connect to the same DTI platform.
Start with the white-label template or connect your existing app. Our initial implementation focuses on ChatGPT, with DTI running as an external platform.
[Your brand] Client Advisor is designed with DTI already connected. Configure the template for your brand and optionally connect your systems using the template’s connectors.
The planned MVP connector set covers five areas:
Start with brand knowledge and a simplified catalogue, then add further integrations as needed. Available data and operational capabilities depend on the integrations enabled.
These connectors are a starting point, not an architectural limit. Additional integrations can be added over time, such as loyalty programmes, consultation booking, service history or product configurators.
Keep your app’s interface, brand-system integrations and commercial workflows. Add the connection to DTI within your own app.
Your app connects to DTI through MCP tools to retrieve expert guidance, access the permitted part of a consumer model and submit event graphs. A server-side SDK is optional. Saved profiles require a sign-in and permission flow.
A graph of events related to the customer’s choice and experience, extracted from the dialogue between the LLM and the customer. The LLM builds the graph according to DTI’s rules, preserving the meaning of events and the relationships between them.
A starting package with an expert consultation strategy and the permitted part of an existing consumer model. The brand also receives consumer choice analytics through an API.
DTI stores permitted event graphs and consumer models. Incoming data is checked for structure, permissions and direct identifiers before core processing and storage. Names, contact details, delivery addresses and payment details are excluded from the consumer model.
A persistent model is created after the customer connects their profile and grants access. The design includes controls to review and correct the profile, authorise connected apps, revoke access and request deletion. Guest recommendations remain available without a saved profile.
Access to a customer’s brand account requires authentication with the brand. Connecting a DTI profile does not replace that authentication or automatically grant access to brand-account data.
Personalisation, consumer choice analytics and service improvement, including training the DTI core, are disclosed purposes, subject to the applicable permissions and legal basis. A brand’s analytics uses permitted information from its own customer interactions. Connecting a consumer profile does not, by itself, grant access to brand systems. Optional integrations are configured separately, within the permissions granted by the brand.
DTI does not store or reconstruct the full conversation.
DTI is grounded in real, individual consultations in a bespoke atelier, transformed into a proprietary methodology for AI-assisted consumer choice.
We are bringing the depth of an expert consultation into a digital product – from understanding the customer to a recommendation they can confidently accept.
DTI’s expert method helps uncover what drives a customer’s choice. Brands gain more than preferences: they understand the reasons behind decisions – why an offer appeals, raises doubts or is rejected.
“It fits the customer” and “the customer is considering buying it” are different outcomes.
Using synthetic consumer models, we explored how customers choose among 680 luxury watches from six brands – revealing shortlist patterns, overlapping product roles and shifts in choice.
Which needs does the assortment meet, where are the gaps, and which suitable products never reach the shortlist?
Why does one product win, another become a compromise, and a third lose? Where do products compete for the same place on a shortlist?
How do changing contexts and priorities reshape the shortlist? Which products retain, lose or gain a place?
Give customers more than a filtered product selection. Help them understand what suits them, compare meaningful alternatives and make confident decisions, with continuity of guidance after purchase.
You do not need an existing app to begin. Use the [Your brand] Client Advisor white-label template with DTI connected, or add DTI to your own app. With the template, you can start with brand knowledge and a simplified catalogue, then add further integrations as needed.
For the pilot, we agree on a customer scenario, the necessary integrations and a primary measure of success: recommendation usefulness, confidence in the choice, task completion or quality of post-purchase support. Where purchase data is available, we compare conversion with a control group.
As sufficient evidence accumulates, the brand gains analytics on the reasons behind choices and rejections. Together, we assess the results, cost of service and next steps.
Discuss a pilot
We are developing Personal Product Advisor, an app in ChatGPT connected to the DTI platform.
Guides you through complex product decisions with professional expertise, shaped around your goals, values, lifestyle, preferences, and individual needs.
Powered by DTI
Description
Guides customers through complex choices like a professional expert: helps them understand what truly suits them and proposes well-founded options based on their goals, values, lifestyle, preferences and individual characteristics.
Designed for choosing products and services in categories that traditionally require expert guidance.
Explore integration, discuss a pilot, request the research or talk about investment. Get in touch with the founder.
Founder and Chief Architect of DTI. Creator of the platform’s methodology, standard and architecture. Combines his own expert consulting practice with the development of personalisation systems.
We use Analytics cookies to understand visits and interactions with this site.
