Sitemap
Next at Chase

We’re telling the story of how we’re becoming one of the industry’s most exciting places to build a career in tech, product management, data & analytics and design.

The AI-Driven Evolution of Enterprise Design

6 min readMay 5, 2025

--

By: Jeff Crossman, Head of Design Systems, Chase

Illustration by John Kim

Enterprises are complex ecosystems, bringing together people, ideas, data, tools and processes. The challenge has always been to align this diversity into a unified entity capable of delivering consistent customer experiences. With the rise of artificial intelligence (AI) and large language models (LLMs), this challenge is evolving. AI is no longer just a tool for specialists; it is democratizing creation across the enterprise, empowering employees at every level to contribute to product development and innovation.

At JPMorganChase, more than 200,000 employees have access to the latest general purpose LLMs, with pilots of more integrated AI expanding using a disciplined approach. These new tools are already accelerating work, elevating skill, and expanding the capabilities of employees.

By lowering barriers to tasks like prototyping, coding and design, AI allows more employees to participate in creative processes. This democratization raises an important question: how can organizations ensure consistency in customer experiences when everyone is a creator? The answer lies in leveraging AI to strengthen the tools of collaboration and cohesion that already exist in the enterprise. Tools like design systems will play a central role in achieving consistency by evolving into smarter, more adaptive frameworks.

Instant Access to Knowledge

One key to consistency is ensuring that all creators have immediate access to the right information and guidance. AI can make design system knowledge available on demand by interrogating documentation, analyzing component code, and referencing support channels to generate suggestions on a range of product design questions with attribution back to source materials.

Today, design systems face limitations in curating content due to the need to maintain a rational information hierarchy. However, by leveraging AI and LLMs, design system teams can now curate and make available a volume of knowledge not previously possible. Expanding efforts to include well-structured research reports, experimentation results, and a deeper understanding of content and accessibility standards will allow design systems powered by LLMs to achieve economies of scale that were previously unattainable with traditional means. This scale will unlock new knowledge while making it instantly accessible to a global workforce.

At Chase, our Manhattan Design System is using Retrieval-Augmented Generation (RAG) to give access to an ever-expanding internal knowledgebase of design, engineering and product knowledge. We are able to provide access to this knowledgebase through the firm’s LLM Suite, in addition to enhancing search on the design system’s documentation site to surface actual answers.

Smarter Component Recommendations

As our design system evolves to accommodate a wider array of products and services, it is inherently becoming more complex. With the introduction of composability into our design system, we can support emerging products and experimentation with a more flexible component architecture, though it introduces new challenges in assembling interfaces correctly.

AI offers a new way to address these challenges by analyzing real-time contexts such as product domain, technical requirements, and even historical usage patterns. AI can suggest the most suitable components for any given task, and more crucially, how to put them together. This reduces errors, accelerates workflows, and ensures that all creators are working from the same playbook.

Additionally, AI can facilitate the migration of older products onto the design system by offering insights into how design system components can be composed to match existing product requirements or suggest updated patterns that could be considered.

The Chase web framework engineering team, led by Samuel Shaw, is doing exactly that. His team created a solution to assist product teams in migrating to our next-generation web architecture. By utilizing custom logic to scan legacy code, this tool can identify dependencies, assemble a very detailed prompt with specific instructions, and employ a RAG-enabled Large Language Model to translate code for teams. This tool also shows promise in recommending components and their configuration from Figma designs.

“AI is no longer just a tool for specialists; it is democratizing creation across the enterprise.”

New Creativity in the Design Process

The traditional design process has, in its most reductive form, two goals:

  • Mitigate bias.
  • Reduce costs associated with building the wrong solution.

With AI, the cost of creativity decreases and, in turn, so does the risk of spending too much to learn a solution doesn’t work. How does the design process change when creating multiple functional prototypes has the same cost as sketching on a piece of paper? When can AI feedback be used as a stand in for real customers? While the design process will still exist, the tools or artifacts employed at each phase may no longer be the traditional ones we are familiar with.

The fluidity of which artifacts are created and when they are used within the design process democratizes creativity. Individuals can choose their preferred creative medium and participate in the design process at any stage. A sketch on the whiteboard, a functioning prototype, a service blueprint — all these things can be created within a meeting to contribute to a discussion. Design systems with assets in multiple modalities enable teams, and AI, to seamlessly transpose ideas across the design and development lifecycle, facilitating new ways to collaborate and be creative.

Building Organizational Memory

Imagine an AI system with improved visibility into what an organization has ever built, tried, succeeded or failed at, along with the reasons why. Such a system would revolutionize the way enterprises approach digital product development. It would allow teams to access a rich repository of knowledge, drawing insights from past projects to inform current and future initiatives. This would not only prevent duplication of effort but also ensure that every decision is grounded in a foundation of accumulated wisdom, where deviations are informed experiments.

While this perfect system is far beyond current capabilities, it does propose a need to ensure that learnings flow more seamlessly across the silos of organizational units than today. Finding what’s succeeding and failing from the edges of the organization, either through manual submissions or with some automation and human curation and making it centrally available could allow large organizations to achieve a level of agility like startups, but with additional benefits of scale and access to proprietary data.

In the future, design systems and the solutions they propose will become more data-driven and trustworthy, attributing solutions back to past successes and evolving with new learnings.

Continuous Quality Assurance

Rather than restricting quality assurance to specific checkpoints, AI enables it to become more ingrained in the process. From design through development to production monitoring, AI can verify adherence to standards like accessibility, spacing, typography and component usage, ensuring human intent is behind decisions and not a misunderstanding of expectations. This proactive approach ensures that quality remains consistent throughout the lifecycle of every product.

Research from the University of California finds that LLMs struggle to generate well-formed accessible code, particularly when it comes to ARIA attributes. Since LLMs are trained from code that often has accessibility issues, it’s conceivable how using them to generate new web code might perpetuate the production of inaccessible code. At Chase, ensuring our experiences are accessible to all customer is of paramount importance. We can overcome this shortcoming by using additional nodes in our LLM chains to reference higher quality, validated Chase production code and additional internal accessibility resources to improve generated code.

Empowering Creators While Maintaining Cohesion

The rise of AI as a creative tool does not mean chaos — it means opportunity. By integrating AI into design systems and workflows, enterprises can empower more of their employees to participate in innovation while maintaining a greater level of consistency than what’s possible today. The enterprise will be able to achieve the agility of organizations that are a fraction of their size while maintaining the benefits and resilience that come with scale.

With AI as a facilitator of human insight, businesses can unlock unprecedented productivity while delivering consistent and impactful experiences for their customers.

Like what you’re reading? Check out our latest opportunities in design and tech.

JPMorgan Chase is an Equal Opportunity Employer, including Disability/Veterans

For Informational/Educational Purposes Only: The opinions expressed in this article may differ from other employees and departments of JPMorgan Chase & Co. Opinions and strategies described may not be appropriate for everyone, and are not intended as specific advice/recommendation for any individual. You should carefully consider your needs and objectives before making any decisions, and consult the appropriate professional(s). Outlooks and past performance are not guarantees of future results.

Any mentions of third-party trademarks, brand names, products and services are for referential purposes only and any mention thereof is not meant to imply any sponsorship, endorsement, or affiliation.

--

--

Next at Chase
Next at Chase

Published in Next at Chase

We’re telling the story of how we’re becoming one of the industry’s most exciting places to build a career in tech, product management, data & analytics and design.

Next at Chase
Next at Chase

Written by Next at Chase

A blog about technology, product, design, data and analytics, and what it takes to build a career at one of banking's most innovative organizations.