AppZen AP Audit
Account auditing made simple.
Remove delays with instantaneous auditing of your accounts.
-
Strategy
UX Strategy
-
Design
UX/UI Design Lead
-
Client
AppZen
In compliance with my confidentiality agreement, some information has been omitted and the design modified.
Coming in two parts, AppZen AP (Accounts Payable) Audit reviews invoices before they are paid, as well as corrals and enforces compliance of their contracts. The program eliminates overcharges, weeds out errors, fraud, and non-compliant spending. It also audits contracts associated with payments and helps customers to negotiate with their vendors.
Eliminates overcharges and duplicate invoices
Validates invoices with PO, contract and external sources
Flags compliance issues
The Challenge
Accountants have large numbers of invoices to process daily, which means sometimes they get processed without validation. The time it would take to do this task manually results in companies losing huge amounts of money each year.
To find out what was needed to solve these issues, we interviewed CFOs, finance controllers, and finance consultant agencies to understand their pain points. After understanding the user issues, I was able to establish my challenges into three clear sections.
Design for the unknown
Imagine trying to design a flying car when you have never even built an engine before. This is what it felt like designing and building a new innovative AI product on a blank canvas.
After detailed market research and validating high-level problem statements, I designed some prototypes. This started discussions with prospective customers, helping to ensure our product would bring value to their businesses.
Our product relies on integration with ERP and Contract Management Systems, so I had to design a way to support dynamically receiving data for beta.
Design for 1.0 innovative product from the ground
I designed concepts for sale demos as well as MVP design for prototype, guiding the way for engineers to turn concepts into a reality. Having this skill, I knew that I would be able to take this challenge in my stride as well. This product has been built in stages, from concept to alpha followed by beta, and then through to MVP for general availability.
Design for AI product with no data
Designing and building a new AI product is very different from traditional tech products. A significant amount of data and AI feedback need to be collected to train AI models, which we need to validate a hypothesis. Without this data, the concept will never eventuate.
A design prototype was presented to some prospective customers to participate in our beta program. They were enthusiastic and approved of the concept, sharing with us the data we needed. This set the wheels in motion, and finally, we were able to kickstart the product.
Validation
Research User Interview
Initially, we interviewed CFOs, finance controllers, and finance consultant agencies to understand their pain points and frauds.
After detailed market and product research, I understood the strength and weakness from competitors. It inspired what value we could bring, position we could make and disrupt the finance industry.
Ideate
After understanding our position and core value, I designed many wireframe and low fidelity mock-up for ideation. It was an interesting stage because I could explore and try creative solutions; at the same time, I need to balance, stay true to the reality that the design is feasible for implementation.
Sales Demo
Our team goal was to find beta customers to validate our proof of concepts. I developed prototypes for sales demos, and it helped to learn our solution would hit the right market.
The prototypes have been through multiple iterations, collaborating with sales team. It has been an interesting and nerve wracking experience because sales team has spent tremendous effort to have one call with Enterprise customers. The few minutes prototype was one of the key elements to grab our opportunities. The customer with purchasing decision was the main persona at this stage.
Finalize Interface and Experience for Beta
After some customers, including fortune 500 companies, participated the beta program, we could gather numerous data to train our AI models and validate our proof of concepts. The experience and interface had gone through multiple major changes, based on feedback, new personas, business decision, data, trade-off, system and technology restriction, manual AI feedback to automation AI feedback stage to what you see now. It’s an incredible journey to design with ambiguity and saw the product evolved so quickly.

The Solution
With the success of our Expense product, we knew we had what it takes to solve these issues as well. We divided and analyzed each problem and addressed them individually. Eventually, these solutions came together and created a way for our clients to turbocharge their AP process.
AppZen Star Match technology validates invoices against contract and purchase orders, as well as other data like logs, messages, badge scans, and sensor data. It only flags invoices that require attention, such as payment term mismatching, pricing misalignment with contracts, suspicious vendors, and duplicate invoices.
The finance industry has been manually trying to avoid error, fraud, and non-compliant spending for years. Finally, with our refined program, we provide a solution our clients can use with confidence.






The Outcome
Initially, AppZen AP attracts a lot of beta customers and validates our hypothesis. By bringing large amounts of data to train our AI model, we have set a great foundation for this product.
Its success perfectly leads into another product in this platform, the next phase, Touchless AP. This is designed to extract invoice data, automate account code mapping, and streamline approval processes. The combination of AP Audit and Touchless AP provides a complete automated Account Payable product.
5%
of suspicious invoices
35%
of suspicious invoice are duplicated charges
38%
of suspicious invoice are price-mismatch with contracts
Millions
dollars of invoices per account flagged as high risk, meaning there were likely errors needing investigation.
Millions
dollars of invoices per account flagged as duplicates of invoices paid in the past 6 months.
The Outcome
With the successful execution of the solutions, we saw some truly incredible results. We continue improving the program, making sure it meets user needs.
70%
reduction of complaints from Red Accounts
2x
faster to review reports
2x
faster to receive reimbursements
24 hours
or less to receive reimbursement in a global enterprise company
>2 months
from research, design, ideate, user interview, develop, and test to launch
We’re changing the way people think of UI/UX creation.
A full-stack structure for our working workflow processes, were from the funny the century initial all the made, have spare to negatives.


Project by the numbers.
App Designs
Creating a full range UI/UX apps accross the whole platform.
API Integrations
Software-enabled platforms to completely change the way of UI/UX iterations.
Markets Campaigns
We have strict intentions to completely change the way of UI/UX creation.
Website & Mobile App Design
0%Awards.
Services.
Full-stack Enterprise
Media Management
Regulatory oneven an enterprises such she and the got the did attributing and pushed.
We currently have 36 active media campaigns across 24 projects.
Google Analytics & Measurement
Spaces of each debt in the digital world can help you with overall simplest authentic;
Get an utilized of structure to tackle complex issues and changes.
UX/UI Design & Website/App Design
A higher spacing and movements through an impactful email campaigns;
Creating a portfolio budgeting in a real time planning and phasing.
Marketing Campaigns & Content Creation
Making a shor time up to date email campaings to achieve a real impact.