ULTIMARII
From 0 to 15 clients: Driving product discovery and UX in AI for regulatory work
PRODUCT DISCOVERY
PRODUCT STRATEGY
AI UX
INFORMATION ARCHITECTURE
LLM DESIGN
ULTIMARII IS AN AI-powered research platform designed to streamline complex document analysis, accelerate knowledge retrieval, and speed up the drafting processes for professionals and large organizations in industries such as energy, infrastructure, and law. The platform enables users to search, analyze, and extract insights from vast document libraries, helping them process compliance-heavy information at a fraction of the time it would normally take.
I was engaged in product discovery and UX strategy, working across multiple sprint cycles to define the user experience, AI interaction model, and system logic that would make this tool not just functional but also intuitive and scalable.
My work spanned:
Sprint 1: Defining the product concept through a storyboard, outlining core functionality such as AI-driven search, document summarization, and user workflows.
Sprint 2: Translating discovery insights into product design, refining the experience based on real-world emerging user needs and nascent industry requirements.
Sprint 3: Refining the UX and UI for MVP launch, problem-solving nesting navigation, cross-project and branching conversations, and simplifying and forecasting for complex and as yet unknown use cases typical in the LLM space.
01
Why now
Professionals and large organizations working with compliance-heavy documents spend significant time and resources manually searching, reviewing, and analyzing dense regulatory and legal text.
The challenge was to build an AI-powered system that could:
Centralize large document repositories, making structured and unstructured data instantly searchable
Leverage AI-driven summarization and analysis to extract key insights in a fraction of the time
Guide users through an AI agent, helping them refine queries, interpret data, and navigate workflows more effectively
Unlike conventional AI research tools, this system had to function at enterprise scale, handling multiple sources of regulatory data with a strong focus on accuracy and traceability.
tl;dr
THE GOAL
Define and design the MVP experience for an AI-powered research platform streamlining complex regulatory workflows
Drive product discovery to refine user needs, LLM interactions and product market fit.
MY APPROACH
Led product discovery, UX research, and MVP UI design to shape the platform's early experience
Defined emerging LLM user interactions and mapped out related user flows
Developed information architecture, user flows, and LLM-driven interaction models
THE RESULT
Helped Ultimarii acquire 15 major clients within its first year
Delivered a clear, simplified user experience for complex large data workflows
Established a structured product foundation, accelerating future iterations and scalability
02
Sprint 1: Laying out the groundwork
Storyboarding the product concept. The first phase focused on defining user workflows through a detailed product storyboard, which became the foundation for early user validation and product direction. It visually mapped out how users would interact with the system, outlining AI-driven search, document summarization, and intelligent recommendations.
The storyboard's primary function was to help present the use case that Ultimarii represented to potential major stakeholders. In this the storyboard acted as a visual proof of concept.
This is a snapshot of the user flow mapping out the initial functions of the product. Some terms were blurred out for confidentiality.
A snapshot of Sprint 1. The images here shows a partial logic flow if the initial product functions, the storyboard as a visual proof of concept and a variety of use case applications, and a concept key visual of the product landing page.
03
Sprint 2: Simplifying large data sets
Two key undertakings were conducted for Sprint 2. One, we needed to understand current proven visualization methods for extrapolating large data sets, refine and simplify the option that was the best fit for Ultimarii's needs, and develop a much more simplified design for version one of the product. Two, LLM interactions are relatively unknown UX space. We needed to define different types of emerging query interactions, predict what the user flows for those interactions will be, and test different iterations of what the interactions can look like in a web-based platform.
Once the core concept was validated, we designed key workflows for:
Smart search with AI-powered retrieval that understands context, not just keywords
Document summarization that broke down long-form reports into concise, structured insights to reduce research time
AI chat interactions that allowed users to ask questions and receive traceable, source-backed responses to ensure trust and transparency
Sprint 2 unpacked current proven way to visualize large data sets, defining emerging types of query interactions unique to LLM platforms, and mapping out the UX and UI for what those interactions can look like.
04
Sprint 3: Problem solving unknown emerging user needs
In preparation for staging, we focused on optimizing information architecture, navigation, and interface responsiveness, ensuring users could interact with the AI-powered system in an efficient manner.
Key refinements included:
Streamlined nesting navigation for managing and retrieving multiple documents efficiently
A structured document viewer for reviewing AI-generated insights with source traceability
Designing the experiences to ensure the AI interactions felt fluid and natural
Snapshot of the summary of work done for sprint 3 where we developed critical functionality for the product. Details are obscured for confidentiality.
05
Where we are now
Through continuous iteration and prototype testing, the AI-powered research system evolved into a functional, enterprise-ready minimum viable product (MVP). By the end of the sprint cycle, the production team now have a clear blueprint to further refine the MVP. This sprint cycle helped accomplish the following:
A fully defined AI-powered research system, validated through early user testing
A scalable UX/UI framework that supports AI-driven search, summarization, and document review
With these improvements, professionals could quickly retrieve insights from thousands of documents, generate structured reports with AI assistance, and streamline workflows that previously took years into a matter of weeks.
WHAT MY COLLEAGUES SAY
Trace is incredibly dedicated and often goes above and beyond to make deliverables or create the best product. She is a very strong leader and has immense skill in leading client workshops, especially in the UX and CDR space.
06
Impact to Ultimarii
Ultimarii made huge strides in just a year, going from an early concept to a fully operational AI-driven research platform that’s already delivering real impact. With 15 enterprise clients including major energy and utility companies and revenue coming in, they’re proving that there’s a massive need for smarter, faster ways to navigate regulatory processes. Their AI-powered system consolidates public and private regulatory data, giving users a clearer path through approvals while cutting down on the complexity and time typically required. They've secured $2 million in funding and plans for a strategic round in 2025 are underway.
A big part of that success comes down to making the platform intuitive to use. From structuring user workflows to refining the way AI-assisted drafting tools function, the UX/UI work played a huge role in making Ultimarii’s offering both powerful and accessible. The product is evolving fast and moving beyond just regulatory approvals into areas like site selection, hearing preparation, and workflow automation. With a strong foundation in place and a clear vision ahead, Ultimarii is set to redefine how organizations navigate regulatory frameworks and they’re doing it with design and usability at the forefront.
High fidelity interface design of the landing page of the web platform from Sprint 2. Actual MVP interface is currently available under early access only.
07
What I learned
When I first started working with Ultimarii, I didn’t fully grasp the weight of who I was working with. Doug, Josh, and the team were approachable, easy to talk to, and genuinely collaborative—just great clients. It wasn’t until later that I truly understood their influence in the regulatory space and the scale of Ultimarii’s ambition.
The design challenges were fascinating. How do you take something as complex as regulatory research, massive datasets, thousands of documents, complex decision-making processes and turn them into an AI-powered tool that simplifies an industry notorious for its complexity?
It wasn’t just about UX, it was about systems thinking at its core. Mapping out the layers of data, understanding how information could be structured for an LLM, and identifying ways to make that intelligence actionable for users required deep design problem-solving. Each sprint was an iteration in refining not just the product’s interface, but how the AI itself would function in a way that made sense for real people.
One of the most rewarding aspects was seeing the shift in perception from Ultimarii’s production team. Initially, there was some reservation, maybe some uncertainty, about whether we could truly help them solve the problems they were facing. By the end of sprint 3, that had flipped. They were leaning on us, deferring to our expertise in product discovery, UX, and problem-solving. It was an unexpected but powerful validation of how UX thinking plays a critical role in AI-driven, highly technical spaces.
This project reinforced my love for solving complex design problems. It proved that thoughtful UX and structured problem-solving don’t just shape a product, they can shift how an entire team approaches its own development.
SHOUTOUTS TO—
Chantelle Little - For trusting me with the opportunity to fly solo on the initial sprint, keep the momentum going and in building a great collaborative partnership with the Ultimarii team.
Josh Malate and Doug Schweitzer - For the easy collaboration, the free flowing whiteboarding sessions, for sharing knowledge about regulatory work and for the trust to get Ultimarii into MVP
Duane Wood and team - Also for the trust! Eventually. I kid. :p Truly thankful for the wonderful collaboration, the deep conversations unpacking the complexity of the design challenges, the grace and patience as we evolved the product into what it is today.
