How AI Can Transform Brex, Ramp, and Expensify for a Smarter Business Future

Recently, I had the chance to experience several expense management platforms like Expensify, Ramp, and Brex (some of which I had used before in previous jobs). This got me thinking about a new dynamic: how can each platform I use daily, whether for expenses, payroll, or organization, improve with AI integration?

In this case, I will focus on three specific platforms:

• Expensify
• Brex
• Ramp

I intentionally leave out platforms like Paylocity, SAP Concur, and Zoho Expenses. Despite being used by large companies, they have too many issues, and my list would be endless. And these three are the ones I've had personal experience with.

Brex, Ramp and Expensify

I used Brex and Ramp regularly in my previous job. Expensify, on the other hand, I was considering because I saw it featured in a few reviews. However, when I searched for it in the App Store, I was surprised to find two different apps from them. Naturally, I downloaded the one marked "new." Still, it left me wondering if one is meant for tracking corporate expenses (both for employees and employers), and the other is for personal use. I eventually found one labeled "expense tracker," which I decided not to bother with.

Before diving into AI features, I wanted to see what other people were saying about these three platforms. Here's what I found:

BREX

Brex's biggest problems, based on the reviews, are:

1. Bad Customer Support - AKA the death sentence of any service-based platform. Many users struggled to get help when something went wrong.

2. Sudden Account Closures - Several businesses had their accounts shut down with little or no warning, causing serious disruptions to their operations. I’m wondering if there was a solid reason.

3. Travel and Payment Failures - Users reported issues with booking flights, where tickets weren’t processed correctly, and payments where money disappeared but wasn’t resolved.

4. Buggy App - The mobile and web apps were described as unreliable, with problems logging in and using security features.

5. Lack of Transparency - Brex made sudden changes, like cutting credit limits or closing accounts, without giving customers clear reasons or enough notice.

RAMP

Based on the reviews, Ramp's biggest issues seem to be:

1. Customer service – Several users complain about inflexible customer service, slow response times, and poor onboarding processes, which leave clients feeling unsupported.

2. Credit limit management – The fluctuating credit limits based on bank balances frustrate many small businesses, especially since these limits can change without notice, leading to operational disruptions.

3. Payment system – The system of taking full payments monthly, rather than allowing for staggered payments like traditional credit cards, is a major concern for small businesses relying on cash flow flexibility.

4. Technical integration issues – Users report problems with integrations, especially with QuickBooks, which is seen as overly complicated and inefficient, requiring extra manual work.

5. Risk management – The reduction of credit lines without clear communication or rationale leaves companies in challenging situations, particularly during times when they need to make purchases or when on business trips.

6. Onboarding experience – Ramp is criticized for providing insufficient guidance and support during the onboarding process, making it harder for companies to use the platform effectively from the start.

EXPENSIFY

The biggest problems with Expensify, based on the reviews, are:

1. Terrible Customer Support - Users consistently mention that it's hard to get help. The chat service isn't useful, and there's no easy way to talk to a real person.

2. Difficult to Cancel - People find it frustrating to cancel their accounts, often getting charged even after trying to cancel.

3. Problems with Onboarding - Hard to soar through your expense management without the proper instructions.

4. Issues with the card expenses and management

5. Hidden Fees - Several users mention that the pricing is misleading, with surprise charges and unclear terms.

6. Outdated Software - While the app was innovative in the past, some feel it hasn’t kept up with the times.

7. Technical Issues - Many users report bugs, duplicated expenses, and problems connecting to accounting systems, attaching files, and navigation.

Now, Here's My Opinion on These Three Platforms

Ramp – Ease of Use

I used Ramp for a long time. The finance person managing the budget could easily create a card whenever we needed to cover differentiated expenses like event costs or ads. The system did read receipts, but most of the time, I still had to manually fill in the description details, which was a bit of a hassle. Occasionally, the card would fail without explanation, which was frustrating. For instance, once I had to request a new card just to purchase a ticket, even though I had enough available credit.

Brex – Customer-Centric

What I loved about Brex is how much they value their customers. They provide excellent support, and you can feel that they genuinely care about their users. While it's true that Ramp offers a better user experience overall, in my opinion, Brex generates more trust. Their interface is also quite solid, but the feeling of being valued as a customer makes Brex stand out.

Expensify – Accessibility

I genuinely appreciate how Expensify’s app is available to everyone, and I can use some features for free. However, based on what I've seen from other features, it feels somewhat basic compared to its competitors. The other platforms seem to have invested much more in user research, resulting in more robust, all-in-one solutions. Expensify may be basic, but to me, it's a "complete basic," meaning it covers the essentials well. However, according to the reviews, it’s the one with the most negative feedback.

Then, because I still believe in the power of brainstorming, I asked my team to think about what other AI features could make for an ideal expense platform or an improved "Brex, Ramp or Expensify" as an exercise. Here’s what we came up with:

1. Automated Receipt Scanning and Data Extraction
(This is generally already a feature, but it’s not very accurate.)

• Use OCR (Optical Character Recognition) to automatically scan and extract data from receipts, invoices, and other documents. AI can categorize expenses based on the extracted text (merchant name, date, amount) and allocate them to the appropriate spending categories.

2. Smart Categorization
(Also already a feature, but can be improved.)

• Implement natural language processing (NLP) to intelligently categorize expenses based on previous spending patterns. The app can learn user preferences over time and automatically adjust categories or recommend appropriate ones based on similar transactions.

3. Improved Onboarding and design with AI:

• AI-Driven Modal Responsiveness
• Dynamic Resizing Detection through the implementation of AI to monitor user behavior during onboarding, especially when they resize the window or change device. By detecting the window resize through machine learning models that analyze patterns in user interactions, the AI can adjust the onboarding modal dynamically without dismissing it.
• Create adaptive UI elements by using AI to create an adaptive user interface (UI) that adjusts modal elements based on the screen size and resolution. For example, the AI can analyze user device data (desktop, mobile, tablet) and predict the optimal way to display the modal without losing content, regardless of window changes.

4. AI-Based User Guidance:

• Proactive User Prompts - The AI can issue smart notifications or prompts when it detects something that may disrupt the onboarding experience. For instance, the AI could prompt the user, "You are changing devices—do you want to continue with what you were doing?" This provides the user with control over whether to pause or continue.
• AI-Powered Error Detection - AI models can be trained to recognize common UX issues like window resizing affecting modals. It can automatically log the resizing event, detect if it causes problems, and adjust UI components in real time to prevent the modal from closing or disappearing.
• Predictive UX Testing - Implement AI to run predictive simulations on different window sizes and device configurations to anticipate. By automating these tests, you can proactively refine the experience to handle various screen sizes and resolutions.
• Personalized Onboarding - User-Centric Onboarding Customization: AI could personalize the onboarding experience based on how individual users interact with the app. For example, if AI notices a pattern where certain users tend to resize their windows frequently, it could proactively optimize the modal for different sizes or suggest a fullscreen onboarding mode, minimizing disruptions.

5. Fraud Detection

• Implement AI-driven fraud detection models to monitor spending patterns and detect anomalies that may indicate fraudulent transactions. These models can alert users in real-time if suspicious activities are detected.

6. AI-Driven Error Prediction and Prevention

• Real-Time Error Monitoring and Prediction - Use AI to analyze user interactions and predict potential errors before they occur. Machine learning models can be trained on historical user data, error logs, and system performance metrics to recognize patterns that lead to errors (such as failed redirects or non-dismissible error messages). The system can then preemptively intervene or suggest fixes in real-time, reducing the likelihood of the error occurring.
• Proactive Error Handling - AI could detect the onset of a potential issue and automatically redirect users to a safe fallback page or suggest alternative steps before the error disrupts the process.

7. AI-Based User Feedback and Guidance

• Contextual Error Messages - AI can personalize and improve the clarity of error messages based on the specific context of the user's actions. Instead of showing a generic error message, the AI can generate custom error notifications that explain the issue in a user-friendly way and suggest possible solutions (e.g., retry, contact support, or auto-resolve).
• Interactive AI Chatbots - Incorporating an AI-powered virtual assistant or chatbot that guides users through resolving the issue. For example, if an error occurs, the AI bot can assist the user in real-time by offering step-by-step instructions to resolve the problem (e.g., refreshing the page, checking internet connection, retrying the action), without the need to contact customer support.

8. Machine Learning for Error Categorization and Resolution

• Error Classification Models - Train machine learning models to categorize the types of errors users encounter on the fraud report page. This would enable the AI system to understand the underlying cause of errors more effectively and provide specific solutions tailored to the type of issue (e.g., missing data, authentication failure, or UI glitch). By recognizing patterns in user behavior and system feedback, AI can apply targeted fixes.
• AI-Based Root Cause Analysis - AI can help developers quickly identify the root cause of the issue by analyzing data logs, user actions, and system states. For example, the AI could highlight the code segment causing the redirect failure or the reason why the error message can't be dismissed, allowing developers to resolve the bug more efficiently.

9. AI-Driven Analytics for Continuous Improvement

• User Behavior Analytics - AI can track how often and under what conditions users encounter this error. By gathering data on user behavior (click paths, page load times, error frequency, etc.), AI can detect patterns that lead to errors and provide insights for preventing them in future updates.

10. User Sentiment Analysis and Feedback

• AI-Driven User Sentiment Analysis - Implement AI to analyze user feedback and sentiment when an error occurs. By detecting negative user sentiment (e.g., frustration from encountering the error), the AI can prioritize fixing the most problematic issues.
• Real-Time Feedback Collection - When an error occurs, AI can collect feedback directly from users (through prompts or surveys) and use natural language processing (NLP) to interpret and categorize the feedback. This helps identify and prioritize issues for future releases.

11. Voice-Activated Expense Entry (A must for Gen Z users!)

• Incorporate voice recognition features where users can verbally record their expenses. AI can process the voice inputs and translate them into categorized entries.

12. Predictive Cash Flow Management

• AI can analyze past and present financial data to predict future cash flow. It can notify users if they are likely to run out of funds based on recurring expenses and income patterns.

13. Expense Trend Analysis

• Provide users with AI-driven insights into spending habits and trends over time. For example, it can notify users of rising expenses in certain categories and suggest ways to reduce unnecessary costs.

14. Personalized Financial Goals (And why not integrate it with a tool like Splitwise that allows me to connect my corporate, personal, and shared expenses with friends?)

• AI can help users set financial goals (e.g., saving for a vacation or an emergency fund) based on their income and spending patterns. The app can also track progress and provide recommendations for staying on track.

15. Expense Splitting and Shared Budget Management

• For users sharing expenses (e.g., roommates or couples), AI can assist in dividing and tracking shared costs, helping maintain transparency and balance in shared budgets.

16. Custom Alerts and Notifications

• AI can customize alerts based on user preferences. For example, users can receive reminders to log expenses after certain transactions or notifications when they are close to exceeding their budget in a particular category.

17. Integration with Financial Institutions

• Use AI to automatically sync with bank accounts and credit cards, categorizing transactions in real-time. The app can use machine learning to identify patterns in transactions that the user might want to track more closely.

18. Goal-Oriented Saving Suggestions

• The app can analyze financial data and provide AI-powered suggestions for reducing expenses and increasing savings, customized to specific user goals.

19. Tax Filing Assistance

• AI can identify expenses that qualify for tax deductions and organize them into tax categories, making it easier for users to prepare for tax season.

20. AI-Powered Content Formatting Consistency

• Smart Content Parsing and Formatting - AI can be trained to detect when content is copied and pasted. By utilizing natural language processing (NLP) models, AI can automatically parse the copied content and ensure that it follows the correct formatting rules.
• AI-Based Content Cleanup - Implement an AI system that cleans up any inconsistencies, such as extra spaces, line breaks, or incorrect HTML tags.

21. AI-Driven Preview Rendering

• Intelligent Content Preview Rendering - AI models can dynamically render previews by interpreting the content and adjusting the UI elements accordingly.

22. Learning from User Edits

• The AI can learn from user edits and adjust its behavior accordingly. Over time, the AI would improve how it handles similar content, reducing the need for manual intervention.

23. Integration with commonly used chat applications

• Furthermore, these platforms should offer integration with Slack, teams or even WhatsApp. For many of us, it's much easier to use WhatsApp for notifications, alerts, and important documents. This way, we could upload receipts and manage expenses directly from WhatsApp chats (or Slack for corporations).

But why stop there?

If a company truly wants to be cost-efficient, expense-tracking apps should act as essential data hubs. Why? Because they can drive business strategy based on expense patterns. For example:

• If a company spends heavily and frequently on employee travel, the app should identify the best times to book flights at lower prices and help configure the company’s travel calendar accordingly.
• If a company has high spending in a specific category, like Ads, the app should help identify ways to optimize those expenses and even partner with providers to secure better rates.
• These platforms should also track data to suggest which contracts and prices a company should negotiate to achieve cost savings.

Conclusion

In essence, a fully AI-enabled expense tracking platform should integrate so deeply with a company that it becomes a core part of its operations. It should be able to securely manage data and help make the company’s spending more efficient.

This kind of transformation would go beyond simple tracking; it would fuse with the company to streamline expenses and improve overall financial strategy. So, my question to the class today, would you be more apt in using Expense Management software if it included some, or all, of the AI features our team of Engineers and Product came up with?