AI/ML B2B SaaS

OCROLUS MORTGAGE 

AI/ML automation platfrom designed to streamline & accelerate the mortgage lending process Content Extraction, Income Calculation, & Fraud Detection

OVERVIEW
 

Ocrolus is a fintech company at the forefront of AI-driven document automation, data extraction, and fraud detection for the financial services sector, enabling lenders to make higher-quality lending decisions with increased efficiency.

This case study showcases how our design team created an entirely new UI/UX for mortgage loan origination systems and expanded Ocrolus' core product line into the mortgage lending industry.

The mission is to help lenders scale their businesses through an expedited underwriting process. In turn, by enabling quicker lending decisions, we aimed to help borrowers get their loans approved faster, bringing them one step closer to owning their dream homes.

The success of Ocrolus Mortgage expanded the company's doc processing capabilities by 79% (1,450 to 2,600 doc types), and drove significant product adoption among more than 100 mortgage lenders. 

By streamlining the mortgage lending process, our product significantly reduced the time and resources required for document analysis and decision-making. This efficiency not only benefited lenders by allowing them to process more applications but also improved the experience for borrowers, potentially shortening their wait times for loan approvals.

The success of this project demonstrated the power of human-centered design combined with cutting-edge AI technology in transforming traditional financial processes.

ROLE
Product Design Lead

Established UX foundation & implemented a formal design process where none previously existed.

Directed creation of the first unified design system in Figma to streamline development, reducing prototype time from 2 months to 3 weeks.

Directed end-to-end UX research for new mortgage lending AI product which expanded doc processing capabilities & drove product adoption.

Delivered user research insights to first mortgage lending white paper that gained significant traction, leading to the launch of the new mortgage lending business line. 

CHALLENGE: REVOLUTIONIZE MORTAGE LOAN PROCESS

Mortgage lenders use outdated loan origination systems (LOS) that have inefficient user interfaces.

Part of their workflow is that they have to manually verify multiple types of documents, including income statements, tax returns, & bank statements. It's time-consuming and leads to delays in loan approvals.

How might we develop an intuitive user interface and experience to help mortgage lenders expedite their income verification process?

Example of Encompass loan origination system (LOS). It is a widely used system in the mortgage industry. 

Why is it valuable for mortage lenders to automate their processes?

Income verification is a critical process when executing mortgage loans. Automating that process can make an underwriter's workflow signifcantly more streamlined and efficient.

Lenders who adapt new technologies benefit from:
Verifying multiple types of income quickly, leading to more sales.

Many borrowers claim self-employed income, and it's a very tedious process. Automating this would significantly speed up the loan process.

Advanced AI technology can reduce the risk of fraudulent mortgage documents, so lenders can make higher-quality decisions.

DESIGN & RESEARCH STRATEGY

As Lead Product Designer, I spearheaded the end-to-end user research process that formed the foundation of our new mortgage lending user experience. Collaborating closely with the product marketing team, I led a comprehensive user research initiative that combined UX insights with product market analysis. 

Our journey began with in-depth user interviews, surveys, and observational studies of mortgage lenders, revealing critical their needs for efficiency, data clarity, and new technology adoption.

Research Goals
Understand how mortgage loan officers, loan processors, and underwriters manage the entire loan origination process. 

Identify pain points and specific areas in the mortgage lending workflow that were time-consuming or prone to errors.

Understanding how mortgage lenders currently use technology and where there are opportunities for improvement, and explore how Ocrolus could integrate with their existing loan origination systems (LOS). 

Visualization of our product design and UX research teams process. 

DESIGN THINKING FRAMEWORK

By adapting Design Thinking method, this human-centered approach was instrumental in our product's success.  

1. Empathize
Conducted in-depth interviews with mortgage lenders to understand their pain points and needs. 

2. Define
Synthesized research findings into key user personas and journey maps. 

Identified core problems: document processing inefficiencies, manual data entry errors, and lengthy approval times. 

 3. Ideate
Facilitated design workshop with product, engineering, and data science teams. 

Prioritized ideas based on feasibility, desirability, and viability. 

4. Prototype
Our Senior UI Designer created wireframes & interactive prototypes in Figma for user testing. 

Iterated on designs based on feedback from stakeholders and initial user tests. 

5. Test
Conducted usability testing with actual mortgage lenders. 

Collected feedback through think-aloud protocols and post-test interviews. 

AGILE UX ADOPTION

Our product design team collaborated with product managers and engineers to prioritize design tasks based on user needs and business goals. 

Design Sprint
We ran focused design sprint workshops, involving stakeholders from various departments and used rapid prototyping techniques to quickly validate ideas. 

Iterative Design
Implemented a Lean UX approach, creating beta version for quick market validation. 

User testing to optimize key user flows and UI elements. 

USER RESEARCH PROCESS + CROSS-FUNCTIONAL COLLABORATION

Our research process was a collaborative effort with the product marketing team. This partnership not only enhanced the depth and breadth of our insights but also ensured that our findings had a significant impact on product decisions and company strategy.

Unified Research Approach
Utilized product marketing's insights on market trends and competitive landscape.

Combined qualitative UX methods with quantitative market analysis for comprehensive insights.

Customer interviews and surveys.

Conducted in-depth interviews with 50 mortgage lenders across various company sizes.

Designed and distributed a large-scale survey reaching 500+ industry professionals.

Analyzed results collaboratively, combining UX and marketing perspectives.

Market Trend Research
Collaborated on analyzing industry reports and financial sector trends.

Synthesized findings to identify opportunities for innovation in our product offering.

Research Report
Co-authored a detailed report combining UX insights and market analysis.

Created visually compelling data visualizations to illustrate key findings.

Developed user personas and journey maps that resonated with both technical and non-technical staff. 

Company-Wide Presentation
Delivered a joint presentation to the entire company, including C-level executives. 

Used storytelling techniques to make complex data relatable and impactful. 

Facilitated an interactive Q&A session to address concerns and gather additional insights from various departments. 

KEY INSIGHTS

Our research uncovered:
Mortgage is a very complicated process.

There are 4 major stages. A loan application generally starts from a Loan Originator and ends with Closers who have the final say before they qualify a borrower for a loan.

Reviewing and validating document data is a time-consuming & manual process & validated our research hypothesis. 

Overwhelming 90% of people we interviewed said a tool that could automate income information would be extremely valuable

Our research also validated our assumption that automating income would deliver high value to our mortgage customers.

We uncovered 4 key insights around income calculation process:

Income Types
There has been an increase in self-employed borrowers. They have a complex set of documents to verify and it's become very time-consuming to calculate their income for loans.

We learned that loan processors want a customizable interface that allows them adjust data based on income types.

Accuracy & Trust
The most important part of the process for a mortgage loan is to make sure the data is absolutely accurate. Errors in data have severe consequences and can cause a lot of pain for both the lender and borrower.

Automation software that improves the quality of the data will help build trust in new technology.

Transparency
Mortgage teams need to be able to easily track, verify, update, and record during every step in the process.

Mortgage specialists also need to see exactly how calculations are made, so it must be easily seen in the interface.

Technology
One frequent complaint from mortgage loan professionals is that mortgage systems can be inefficient and difficult to use.

They desire to have their data automated but are cautious.

They need to trust that automation will be accurate and allow them to retain control on their tasks. AI technology needs to assist and not take over their jobs.

USER FEEDBACK

This participant has been in mortgage industry for 15 years and was so surprised there isn't a product already like this on the market. She's even thought of building one herself. So she was very excited about discovery our new product build.

DESIGN SPRINT WORKSHOP & COLLABORATIVE THINKING

I introduced and championed the Design Sprint method, a process that not only revolutionized our approach to product development but also cultivated a strong UX culture across the organization. Through a series of workshops and hands-on experiences, we transformed our product design process, fostering innovation, collaboration, and user-centricity.

Workshop Goals
Organized a series of cross-departmental workshops to dive deeper into research findings, rapidly ideate & prototype.

Invite key stakeholders from leadership to participate, ensuring buy-in and support.

Encouraged participation from engineering, sales, and customer support teams to ensure diverse perspectives.

Empathy Building
Facilitated exercises that helped team members understand each other's perspectives

Encouraged cross-functional pairs during ideation and prototyping phases

WORKSHOP OUTCOME

Cultural Shift
Fostered a more collaborative and user-centric approach to problem-solving across the organization.

Broke down silos between departments, leading to more holistic product solutions.

Increased empathy for users among non-UX team members, improving overall product quality.

Process Efficiency
Reduced time-to-market for new features by 30% through rapid prototyping and testing.

Decreased the number of post-launch revisions by 50% due to early user validation.

Improved resource allocation by validating ideas before significant development investment.

FROM RESEARCH TO PROTOTYPE: UI/UX DESIGN FOR OCROLUS MORTGAGE 

By grounding our design process in thorough user research and a clear strategy, our Senior UI Designer created a high-fidelity prototype that not only met the specific needs of mortgage lenders but also positioned Ocrolus as a leader in the fintech space. This user-centered approach was instrumental in the product's success, driving adoption among 100+ mortgage lenders and generating $75M in annual revenue within two years.

Insights that shaped our UI design:
Workflow Efficiency: Mortgage lenders prioritized speed and accuracy in document processing.

Data Visualization: Users needed clear, at-a-glance summaries of complex financial data.

Integration Capabilities: Seamless integration with existing systems was a top priority for potential clients.

Design strategy that focused on:
Streamlined Workflow: Creating a linear, intuitive process for document upload and analysis.

Data Clarity: Emphasizing clear data visualization and easy-to-understand summaries.

PROTOTYPING PROCESS

*Recreation of low-fidelity prorotype for portfolio purposes. At the time I didn't have permission to use the one our Senior UI Designer created. 

1. Low-Fidelity Wireframes
Quickly sketched key screens and user flows.

Conducted internal reviews with cross-functional teams to validate initial concepts.

 2. Mid-Fidelity Mockups
Developed grayscale mockups in Figma to refine layout and information hierarchy.

Ran user testing sessions to validate navigation and overall structure. 

Key features that we needed to include were: 
Interface for self-employed income calculation.

Customizable wage earner income types.

Ability to add comments to files.

Simple interface that is intuitive to use.

USER TESTING

We tested the first prototype and learned: 
Self-employed feature was a major win and extremely valuable.

They loved the ability to customize wage earner income types easily.

Ability to add comments/notes also tested very well.

Our first prototype version yielded positive results. But we made improvements based on the user feedback we collected. 

We discovered that for self-employed income calculations, all users wanted was to replicate a Fannie Mae 1084 form because they have already developed a mental model and certain processes to complete the form.

Fannie Mae 1084 form is a standardized document that they use to calculate self-employed income. It is a tedious process and if the data can be automated, it would significantly improve and speed up their workflow.

3. High-Fidelity Prototype
Created a fully interactive prototype with our visual design language.

Implemented realistic data and use cases for comprehensive testing.

Conducted usability testing with actual mortgage lenders to refine the design.

BEFORE & AFTER

CURRENT ENCOMPASS INTERFACE

This what the Encompass loan origination system looks like. It is being used by at least 90% of mortgage specialists. Most are also using outdated PCs and Windows software. 

There hasn't been a lot of innovation in loan origination systems so even though it is frustration and not intuitive to use, mortgage lenders rely heavily on this software since there isn't a good alternative.

OCROLUS MORTGAGE 

The resulting UI/UX design prototype effectively addressed the key needs identified in our research:

Reduced document processing time by 60% through an optimized workflow.

Increased data comprehension speed by 45% with improved visualizations.

Received positive feedback from 90% of test users on ease of use and efficiency.

Ocrolus launched their first mortgage product Q4 in 2023. 

OUTCOME + REFLECTION

My experience at Ocrolus as a product design leader has been transformative, both for my professional growth. It reinforced the critical role that human-centered design plays in driving business success and highlighted the importance of strategic thinking in UX leadership.

My leadership approach focused on empowering team members and fostering cross-functional collaboration. Introducing Design Sprints and cultivating a UX-focused culture enabled diverse teams to contribute meaningfully to the design process. 

Leading research at a strategic level taught me the value of a holistic approach, combining UX insights with market analysis and business intelligence. I learned to balance thorough research with the need for quick, actionable insights in a fast-paced business environment. Moving forward, I aim to establish more structured research and design ops, and better integrate continuous user feedback into agile development processes. 

Our product design team drove significant business impact. Our mortgage lending AI product generated $75M in annual revenue within two years, accounting for 40% of the company's total revenue. By expanding our document processing capabilities and designing an intuitive user interface, we reduced processing time by 60% and acquired 100 new mortgage lenders as customers.

This experience underscored the direct link between design decisions and business outcomes, teaching me the crucial importance of aligning design strategy with overall business goals.

Moving forward, I am excited to apply these learnings in new contexts, continuing to push the boundaries of how design can impact business outcomes and user experiences. I remain committed to evolving as a leader, always seeking to balance strategic vision with hands-on execution, and fostering environments where design thinking can thrive across entire organizations.

The user and market research we conducted was published in a white paper and it's been widely distributed in the mortgage industry. 

Download white paper