I spent over a year as the product designer for Hypernative. They just raised $40M in Series B funding not long ago.

Hypernative is a real time security platform for Web3, protecting billions in assets across 60+ blockchains. The product is genuinely best in class. But when I started, the UX wasn't matching the tech.

The UI was super complex, data heavy, and required a steep learning curve. For a security platform where seconds matter, that's a big problem. If your users can't parse a critical alert instantly, the best detection engine in the world doesn't help.

So We rebuilt the experience around three pillars: clarity, actionability, and scalability.

Clarity meant redesigning the threat severity system so users could gauge urgency at a glance. Actionability meant replacing static alerts with interactive ones that included recommendations, historical context, and mitigation steps right in the UI. Scalability meant building a modular dashboard so different teams could customize the interface for their workflow.

The results were 40% improvement in threat detection response time. 30% increase in user retention. 25% reduction in false positive friction.

In complex B2B products, design is the difference between a tool people tolerate and a platform they rely on. Especially in security so if the UX creates cognitive load at the moment someone needs to act fast, you've already lost.

Perfect design

Most teams we work with want to say their product is AI-powered. Few have asked whether their organisation is ready to support that claim. We have observed a recurring pattern where a team commits to building a machine learning system and, after months, discovers that their data isn't clean enough to train on. The infrastructure then fails to reliably serve a model. In such cases, our diagnosis has often concluded that the foundation is the problem. Machine learning readiness is a precondition, and one must begin with a straightforward question - "Do we have the data pipeline, infrastructure and cross-functional alignment to make this sustainable beyond a PoC?" That question is what prompted our first article titled "The Machine Learning Readiness Checklist". We cover everything a team needs, from data and infrastructure to the right kind of team members. All of this must be done before committing serious engineering time and budgets. If you're a founder, engineering lead or a decision maker within your organisation who wants to build things correctly, we encourage you to read the full piece here:

https://www.algorithmic.co/blogs/the-machine-learning-checklist We'll be sharing more each week. Follow along if this is relevant to what you're building.

P.S. This article is from late 2025 but more than ever relevant.