After Years Designing SaaS Apps, Jeremy Blaze Thinks AI Is Going To Kill Most

Jeremy Blaze, one of tech's most perceptive design thinkers, suggests that the sleek dashboards that have defined software for a decade are headed for extinction. At a time when most designers are perfecting traditional interfaces, he is already mapping their replacement.

Software design has been dominated by the SaaS model— dashboard-heavy, admin-focused applications packed with tables, forms, and validation flows for years. With AI becoming more sophisticated, this paradigm is about to change. These are theoretical musings and conclusions drawn from his extensive hands-on experience. As the founder of Never Before Seen, a product design agency that has become the secret weapon for ambitious startups, he has architected interfaces for over a dozen vertical SaaS companies. This extensive exposure has given him a panoramic view of current software design patterns and limitations.

His portfolio showcases standout projects, including Ultralight's medtech product and compliance hub, Programa's design workflow tools and Astra's security platform. This specialised approach has positioned him as a key leader in creating tailored software solutions across multiple sectors.

Beyond Dashboards: The Agent Revolution

From his extensive work with SaaS interfaces, Blaze identifies a fundamental problem with traditional applications. 'They are expensive to build because every interaction requires a bespoke interface. These custom views and wizards demand significant development resources,' he explains. 'So rather than forcing humans to learn complex logic, next-generation systems invert the relationship, with AI adapting to human communication patterns.'

While traditional SaaS interfaces will not disappear overnight, he envisions their evolution into more adaptive, contextual, and conversational systems. Designers will shift from rigid wizards to flexible environments that respond to user needs in real time.

Collaborative Workspaces

Blaze identifies one of the most profound shifts toward collaborative AI workspaces that blur the boundary between human and machine contributions. Traditional systems rely on static fields while emerging design patterns create shared environments where AI is a persistent assistant.

Blaze also points to pioneering examples such as ChatGPT's Canvas feature and Lovable's Developer Copilot interface, which combines natural language commands with visual editing to build applications without manual coding. The interface enables developers to describe desired changes and see them implemented instantly, featuring a unique 'Select & Edit' function for modifying elements through verbal requests.

Behind the scenes, AI agents can make these same changes autonomously, working with the codebase directly. Unlike traditional coding assistants, Lovable is a complete development partner capable of constructing entire applications from prompts.

Moreover, as AI assumes greater responsibility, Blaze identifies trust as a critical foundation. Systems must communicate what is happening and provide oversight. He advocates for interfaces highlighting AI-generated changes. Deepseek exemplifies how transparency can help users learn from AI reasoning patterns while algorithms adapt to feedback.

It highlights its approach at the model level by explaining how tasks are handled. Similarly, ChatGPT Operator's 'confirmatory interface' demonstrates this principle by pausing before executing critical actions—like purchases or sending emails—to request explicit user confirmation. These mechanisms represent key steps in fostering trust between users and AI systems.

The Future of Interface Design

As software evolves from passive tools to proactive agents, Blaze envisions a major shift in managing tasks. Unlike traditional systems where users organise all work, Blaze sees humans and AI agents working from the same task list as collaborative partners. Users will assign tasks to AI suited for automation, while agents will delegate tasks requiring human judgment back to users. This bidirectional workflow allows each to leverage their unique strengths. Blaze anticipates systems with clear visibility of completed actions by both parties. These capabilities require careful balance, ensuring the relationship remains productive while preserving user control.

The implications of Blaze's vision extend beyond aesthetic changes. These suggest a fundamental reconsideration of how humans and technology collaborate. As AI grows more capable, he anticipates decreased reliance on manually driven CRUD apps and increased emphasis on dynamic, AI-assisted workflows.

For designers, key questions have evolved from 'How do we organise this information?' to 'How do we ensure the AI surfaces the best information?' and 'How do we create interfaces that feel like working with a competent assistant?' These challenges require a deeper understanding of psychology and language and trust in intelligent systems.

The future Blaze envisions is not about refined dashboards. Instead, it focuses on designing environments where humans and AI collaborate seamlessly, each contributing unique abilities.

Through pioneering work at Never Before Seen, Blaze is theorising and actively creating this future of human-AI collaboration. Blaze's product, Blair, demonstrates how radically different approaches can yield more intuitive software solutions that transform how we interact with technology. This practical implementation shows how innovative interface design and thoughtful AI integration can bridge gaps between traditional software models and new paradigms.

Companies can evolve or become obsolete as traditional SaaS interfaces face increasing pressure. Blaze and his team at Never Before Seen guide forward-thinking organisations through this transition— reimagining interfaces and the fundamental relationship between humans and software.