Databolt is a vaultless tokenization solution that integrates with internal apps and tools like Snowflake and Databricks to secure sensitive data with tokens.

As a new platform, it lacked metrics for performance, adoption, and ROI; clients migrating large datasets couldn't tell if activity drops meant completion or failure.

I led the 0→1 analytics design—defining "visibility," launching an MVP platform health dashboard, and expanding to granular error diagnostics, coverage monitoring, and anomaly detection to build trust.

Company
Capital One

Team
1 PM and 4 Engineers

Timeline
MVP: 8 weeksMVP+: 3 weeks

Takeaways
Product Influence · Data Visualization · UX Strategy

The Problem

Databolt was a brand-new tokenization platform with no metrics to show its performance, adoption, or value relative to cost. B2B clients migrating millions of sensitive records had no visibility into whether tokenization was working. When activity dropped, they couldn’t tell if a job had finished or failed, and they had no data to justify investing in the platform.

The Solution

As lead designer, I took Databolt’s first analytics experience from research to launch. My research shaped the PRD, and I designed a dashboard that helped clients monitor platform health and catch issues early. Since launch, Databolt has signed 2 B2B contracts, with Token Analytics and Audit named as the deciding experiences. Nine months later, my end-to-end evaluation made the case for the next phase: data coverage, traceable errors, and anomaly detection.

Final Outcome

Our year-long goal was to increase clinician sign-on, increase order completion, and update our outdated Sales rep mobile app. The project has three parts: research that results in a pivot of direction, MVP design, and MVP+ fast follow.

The Final MVP Design

Creating A Callout For Clinician Signature Needed

To eliminate searching for clinician signatures, our PM proposed digital signature requests to boost sign-offs and sales rep engagement by 3k monthly. I designed the MVP using our design system to improve order conversion.

The Final MVP Design

Updating the visual hierarchy for readability and fixing bugs

Researching or restructuring the order detail page was out of scope. However, I did update the spacing, gave the page more visual hierarchy, and fixed the button bug.

The Final MVP+ Design

Filter Orders to Find Their Most Crucial Needs

Sales Reps want to find orders with missing documents that need urgent turnaround. Currently, the user needs to click each order to find this information.

The Final MVP+ Design

Searching Orders by Facilities or Patients Based on Sales Rep’s Mental Modal

I emphasized the need for a solution to decrease sales rep turnover. I improved the app's usability, enabling sales reps to prioritize work and patient follow-ups during facility visits. This resulted in faster order completion and increased app usage.

Outcome

We are still in the early months of our first release; however, we have seen a 20% increase in sales rep app. usage from our suppliers, such as Lincare and Adapt. Also, we see a 25% increase in clinician sign-on. We are in the process of releasing the filter experience.

Research

Timeline
2.5 weeks

My Contribution
User Reseach and Analysis

Research: Problem

Ownership of recurring orders was disputed. Some advocated sales rep management based on supplier feedback, while others cited insufficient incentives for reps to follow up consistently.

Users needed a way to understand token generation, performance, and the platform's operational and financial value. Without these insights, stakeholders lacked visibility into platform health, system errors, and return on investment. A team migrating sensitive data couldn’t tell whether a drop in activity meant the job had finished or failed.

Research: Outcome‍ ‍

My research became the foundation of the Token Analytics PRD. I led the research that narrowed it down: defining what “visibility” should mean for the MVP and identifying what data engineering could actually provide.

Through persona workshops, a backend data alignment session, and interviews with Turning (Capital One’s internal tokenization team), I defined the user problems, the decision-making needs, and the specific metrics the MVP would surface. I wrote these findings directly into the PRD’s user needs, requirements, and success metrics, setting the scope for what Product and Engineering built

01: Interviewed the Turning team

Through interviews with an internal team, I learned how similar teams read token analytics: which signals they prioritize, how they structure information, and how they spot anomalies.

02: Facilitated a 90-minute persona workshop

I designed and facilitated a workshop with stakeholders to align on the business problem, target users, and dashboard goals. This grounded the experience in real operational workflows rather than assumptions.

Research Findings

Based on stakeholder and internal team feedback, I grouped my discoveries into three core user needs. Our findings gave stakeholders a shared way to prioritize capabilities against the platform’s long-term vision.

Discover Data

 Explore token behavior and monitor risk signals.

Analyze Data

 See when, where, who, and how data is tokenized or de-tokenized.

Manage Data

 Report and audit platform value to justify investment.

03: Feasubility Workshop

I led a feasibility workshop with the PM and Engineering lead to map user needs against the data the backend actually captured. Together, we prioritized the Analyze pillar for the MVP.

04: Synthesized Findings to Product Requirement Document (PRD)

I synthesized findings from workshops and stakeholder interviews to influence the Product Requirements Document (PRD). I recommended the specific metrics and data signals to surface in the MVP analytics dashboard, ensuring the experience would provide meaningful insights into platform performance while establishing a scalable foundation for future analytics capabilities.

Research Outcome

I co-wrote the final PRD with my PM, turning my research into the metrics and data signals the MVP would surface. This made sure the team built from real user needs rather than assumptions, and it set a scalable foundation for future analytics.

MVP Design

Timeline
4 weeks

My Contribution
Interaction Design

MVP: Problem

As a new 0→1 tokenization platform, we needed to show prospective users our performance, adoption, and value relative to cost.

The product offered no real-time visibility into how tokenization worked across systems or when issues escalated. Without analytics, teams couldn’t monitor performance, detect anomalies, or evaluate platform effectiveness.

MVP: Outcome

Beyond designing the primary tokenization analytics experience, I spearheaded error visibility and threshold tracking.

Despite leadership concerns about displaying errors on a new platform, I advocated for a success-rate indicator to maintain user trust, enabling fast detection of performance issues.

Since real-time alerts were technically unfeasible for the MVP, I added a 5-second threshold marker on the chart for users to spot latency risks.

02: Advocated for a single-page dashboard.

After exploring multiple layouts, I designed one view showing tokenization speed, where it occurs, and whether errors are present. Users could check platform health at a glance without jumping between screens to piece together what was happening.

01: Prioritized The Most Valuable Metrics.

Early concepts included unique token counts and detailed error metrics. I worked with PMs to pare these down to what onboarding teams needed most, keeping the dashboard focused so key signals wouldn’t get buried in noise.

Set time ranges around real workflows

I pushed for 24-hour and 7-day views instead of the short default intervals, with summary data at the top before the detailed graphs. Migrations run over hours or days, so users could get the big picture first, then follow a full job from start to finish and confirm it was completed.

MVP: Final Design

Matched visualizations to the data

A line graph shows speed over time, and bar charts compare tokenization vs. de-tokenization and field types. Users could spot slowdowns and unusual activity at a glance, without digging through raw logs.

MVP: Final Design

MVP: Final Design

Designed around a technical constraint

Research showed that latency over 5 seconds signals trouble, but real-time alerts weren’t feasible for the MVP. I added a 5-second threshold marker to the speed chart, so users knew right away when performance crossed into risk.

MVP: Final Design

Advocated for showing errors.

Leadership worried that surfacing failures would hurt confidence in a new platform. I argued that hidden errors would cost us users’ trust. The success-rate indicator let users catch problems early instead of finding them after the fact.

Outcome

After launch, Token Analytics and Audit were the two experiences that won over prospects and helped Databolt sign its first two major contracts. I designed a focused, single-page dashboard that turned complex tokenization activity into clear signals. B2B clients could check platform health at a glance, follow migrations from start to finish, and catch performance issues early.

MVP+

Timeline
3 weeks

Artifacts
Product Roadmap
Final MVP+ Design

MVP+: Problem

Our team faced two critical challenges with the Sales Rep mobile app: developing a feature for quickly identifying crucial orders and creating a comprehensive roadmap to improve the app experience.

The MVP design successfully solved the business goal but did not help the sales reps quickly find their crucial needs. If the team wanted to revive the app, we needed to improve its discoverability.

MVP+: Outcome

I implemented filters and inline searches for urgent order identification based on internal workshop findings and usability tests. Preliminary feedback from key clients like Linden indicated significant improvements in sales rep efficiency.

The data from the usability test gave us insight into how the sales reps search for orders. I wanted to apply it in the fast-follow design.

03: MVP+ Design

Data from the previous usability test showed a high demand for filters to find the sales reps’ most crucial needs.

With the data, I convinced the PMs that we needed a fast follow-up. The PMs and I created a roadmap for the sales rep app. I worked with the engineers to understand the feasibility of the new feature.

01: MVP+ Design

I organized a workshop with sales experts, PMs, and engineers to strategize the future of the Sales Rep mobile app experience.

I planned a workshop with two PMs, seven internal staff who support the Sales Reps, and two Lead engineers for us to discuss the key pain points, needs of the Sales Reps, and the gaps within our current experience.

02: MVP+ Design

We organized the features for our sales reps from the least to the most important.

We organized the different features by impact and feasibility using the matrix framing exercise (time x impact). Ultimately, we listed the filters from least to most important.

More Usability Testing Findings

With the engineers’ feedback on feasibility, I recommended these features from the prior usability test to meet the users’ need to discover their crucial orders and prioritize their work. These additional findings help me convince my PM to do a fast-follow.

(MVP+) Feature One: Search Filter

Four of five sales reps stated they wanted to filter by facilities and patients, saying that it helps them prioritize their work for the day.

(MVP+) Feature Two: Other Filters

Sales reps thought missing documentation and other attributes were equally important as a clinician’s signature request.

(MVP+) Feature Three: Helpdesk

Sales Reps need to update orders. However, they all prefer to contact a Parachute Assistant to make those changes.

MVP+: Final Design

Search Filter

The search filter was in-page, allowing the sales rep to prioritize their work during facility visits and follow up with patients, increasing sales by search by facility or patient. Working with engineers, we decided not to have an auto-populate experience for the first release.

MVP+: Final Design

Additional Filters

The additional filter helped prioritize outstanding orders, not solely outstanding signature requests, so I added the “outstanding signature request” to the filters as “actions needed.” The next step is to explore a saved filter experience after the fast-follow.

Helpdesk Feature

MVP+: Final Design

For MVP, the Helpdesk future was out of scope. However, I created an experience that allows the sales rep. to email support if any changes are needed, like the clinician’s contact information.

Outcome

When I left, we were still collecting data, but top clients like Linden reported significant improvements in sales reps' efficiency. The inline search and filters helped reps find their most crucial orders faster, increasing their likelihood of using the mobile app more frequently. Additionally, the enhanced experience led to increased order completion rates, which could reduce facility and clinician churn.

Lessons Learned

  • Balancing Business Goals with Users' Needs

    As designers, we must evaluate and map out ways the business goals meet the user needs. I collaborated with the PM to balance business objectives and essential user needs, reducing facility churn and increasing order deliveries.

  • Working within Tight Deadlines To Meet Goals

    Over two years of neglected mobile app research resulted in bugs and sales rep abandonment. With no data and a tight timeline, I had to conduct research and usability testing based on assumptions quickly.

  • Using Usability Testing to Pivot Direction

    As a strong proponent of usability testing, I involved my PMs and engineers in this new process. I educated them on its significance by including them in test sessions, collaborating on research scripts, and explaining my analysis process.