Why We Built UserFlux to Lead This Transformation
Over the past two posts, we’ve made the case that the web is shifting from static browsing to dynamic, agent-driven experiences — and that customer data platforms are uniquely positioned to power them.
This final post is about UserFlux, and why we’ve spent the last few years building the platform that makes this real.
Built from the ground up
UserFlux is a customer data platform. We built it with the same foundational capabilities that any serious CDP requires: product analytics, identity resolution, user profiling and enrichment, marketing automation, and direct SQL and API access to the underlying data warehouse.
These aren’t side features. They’re the infrastructure that everything else depends on.
Over the past year, we extended the platform into product search and personalisation. Customers can ingest their full product catalogue into UserFlux, deploy configurable ML models with a single click, and serve personalised search results and recommendations to their end users — all powered by the behavioural and identity data the platform already holds.
The results have been strong. Search and personalisation work dramatically better when they’re built on top of rich, first-party customer context — not bolted on as a separate system.
Introducing the UserFlux AI SDK
Now we’re packaging all of this into something new: the UserFlux AI SDK.
With a few lines of code, any business can embed an entirely new kind of discovery experience directly into their product. The SDK exposes three core capabilities:
Search. Fast, personalised product search powered by UserFlux’s indexing engine and ML models. Results are ranked based on who the user is, not just what they typed.
Personalisation. Real-time product recommendations driven by the full context of a user’s behaviour, preferences, and profile — served through a simple API call.
Agentic discovery. This is the new capability. A generative, goal-oriented UI that helps users accomplish tasks — not just find products. Powered by LLMs, grounded in your first-party product and customer data, and rendered directly in your product.
The agentic function understands what the user is trying to do, asks the right follow-up questions, and assembles a complete solution from your catalogue — all within a minute.
Why UserFlux, specifically
When an agent powered by UserFlux has a conversation with a user, it has access to everything: who the user is, what they’ve done before, what products exist, what’s in stock, what’s relevant to this specific person in this specific moment. That context is what turns a generic chatbot into a genuinely useful product experience.
We didn’t build a framework and hope businesses would bring their own data. We built the data platform first, and now we’re putting the right interface on top of it.
It’s even in the name. UserfluX — UX — User Experience.
What’s next
The UserFlux SDK is available now in early access. If you’re a product or engineering team looking to offer your users a fundamentally better way to discover and buy your products, we’d love to talk.
The web is becoming more conversational, more personalised, and more agentic. We built UserFlux to make sure your product leads that shift.
This is the final post in a three-part series on the future of product experiences.