
From Manual Entry to AI-Powered Expense Management
How might we streamline receipt capture and expense tracking to reduce manual effort and improve accuracy across user roles?
Rent Manager (London Computer Systems)
Role: UX Designer — Collaborated with product analysts and developers
SkillsUX Design
Interaction Design
Cross-Platform Experience Design
The Challenge
Background
Rent Manager is a property management platform that includes accounting tools for tracking company expenses. Users frequently handle high volumes of receipts—from maintenance purchases to team expenses and reimbursements.
Objective
Reduce manual effort in managing company transactions and reimbursements by designing an AI-enhanced receipt processing experience.
Before this feature, receipt management was entirely manual.
Key Problems
Opportunity
Create a centralized, intuitive system for capturing, processing, and managing receipts—while leveraging AI to reduce user effort.
Goals
Understanding the Users
To ensure the solution fit into real workflows, I conducted user interviews focused on how receipts are currently handled across roles.
Key Users

Mike - Expense Manager
Oversees expenses and approvals
Needs

Tom - Field Technician
Captures receipts on the go
Needs

Annie - Administrator
Processes and records transactions
Needs
Key Insight
Receipt management isn’t just a single task—it’s a multi-step workflow across multiple users. This revealed a specific design need.
The experience had to support both quick capture in the field and efficient processing in the office, without creating extra work at either stage.
Design Approach
I broke the experience into two core workflows:
This allowed me to design for different contexts while maintaining a connected system.
Designing Receipt Capture (Mobile)
The first challenge was simplifying how users capture receipts in real time. The image capture process we needed to explore had three distinct steps.
Capture single or multiple images
→
Group images into one receipt
→
Decide what to keep, retake, or discard

Iteration on this small but crucial step of the experience started with wireframing. Refining this process included screens for a clear flow for capturing and confirming receipt images, the ability to combine multiple images into a single receipt, and lightweight data input to complement AI extraction
Result
A fast, flexible capture experience that works in real-world conditions (in-store, on the move).
Designing Review & Transaction Creation (Web)
The second phase focused on how receipts are processed and turned into transactions.
There were three main design challenges.
Display of multiple images per receipt
→
Group receipts into one transaction
→
Balance dense information with clarity
I iterated on:

Iteration on this small but crucial step of the experience started with wireframing. Refining this process included screens for a clear flow for capturing and confirming receipt images, the ability to combine multiple images into a single receipt, and lightweight data input to complement AI extraction
Result
A structured, scannable interface that allows users to quickly review, validate, and finalize transactions.
Outcome & Impact
This feature solves many of Rent Manager users’ needs. It eliminates the need for third-party receipt tools, significantly reduces manual data entry, improves speed and accuracy of transaction processing, and creates a seamless workflow from receipt capture to final transaction.

From Manual Entry to AI-Powered Expense Management
How might we streamline receipt capture and expense tracking to reduce manual effort and improve accuracy across user roles?
Rent Manager (London Computer Systems)
Role: UX Designer — Collaborated with product analysts and developers
SkillsUX Design
Interaction Design
Cross-Platform Experience Design
The Challenge
Background
Rent Manager is a property management platform that includes accounting tools for tracking company expenses. Users frequently handle high volumes of receipts—from maintenance purchases to team expenses and reimbursements.
Objective
Reduce manual effort in managing company transactions and reimbursements by designing an AI-enhanced receipt processing experience.
Before this feature, receipt management was entirely manual.
Key Problems
Opportunity
Create a centralized, intuitive system for capturing, processing, and managing receipts—while leveraging AI to reduce user effort.
Goals
Understanding the Users
To ensure the solution fit into real workflows, I conducted user interviews focused on how receipts are currently handled across roles.
Key Users

Mike - Expense Manager
Oversees expenses and approvals
Needs

Tom - Field Technician
Captures receipts on the go
Needs

Annie - Administrator
Processes and records transactions
Needs
Key Insight
Receipt management isn’t just a single task—it’s a multi-step workflow across multiple users. This revealed a specific design need.
The experience had to support both quick capture in the field and efficient processing in the office, without creating extra work at either stage.
Design Approach
I broke the experience into two core workflows:
This allowed me to design for different contexts while maintaining a connected system.
Designing Receipt Capture (Mobile)
The first challenge was simplifying how users capture receipts in real time. The image capture process we needed to explore had three distinct steps.
Capture single or multiple images
→
Group images into one receipt
→
Decide what to keep, retake, or discard

Iteration on this small but crucial step of the experience started with wireframing. Refining this process included screens for a clear flow for capturing and confirming receipt images, the ability to combine multiple images into a single receipt, and lightweight data input to complement AI extraction
Result
A fast, flexible capture experience that works in real-world conditions (in-store, on the move).
Designing Review & Transaction Creation (Web)
The second phase focused on how receipts are processed and turned into transactions.
There were three main design challenges.
Display of multiple images per receipt
→
Group receipts into one transaction
→
Balance dense information with clarity
I iterated on:

Iteration on this small but crucial step of the experience started with wireframing. Refining this process included screens for a clear flow for capturing and confirming receipt images, the ability to combine multiple images into a single receipt, and lightweight data input to complement AI extraction
Result
A structured, scannable interface that allows users to quickly review, validate, and finalize transactions.
Outcome & Impact
This feature solves many of Rent Manager users’ needs. It eliminates the need for third-party receipt tools, significantly reduces manual data entry, improves speed and accuracy of transaction processing, and creates a seamless workflow from receipt capture to final transaction.