Midday

Designing an AI Assistant that simplifies financial data.

Role

Product Design Lead

Industry

Technology

Read Time

5 minutes

Turning Complex Financial Insights Into Clear, Actionable Conversations

When I started working on this project for Midday, the main objective was to design an AI-powered assistant that made it easy for users to understand and interact with their financial data. Midday is a platform that provides businesses with deep financial insights, but many users struggled to turn that data into decisions. They needed a more intuitive way to explore their information without digging through multiple dashboards or manually interpreting charts. My goal was to design an experience that transformed complex data into a conversation that anyone could understand.

Designing Data Conversations Instead of Dashboards

The first major shift I focused on was changing the way users interacted with information. Instead of relying on traditional reports and static dashboards, I designed an experience where users could simply ask questions in plain language and receive meaningful answers. This meant building an interface that felt conversational and natural while still being powerful enough to deliver real insight.

Users could type questions like “What was my revenue growth last quarter?” or “How does my spending compare to last year?” and the AI would respond with clear, contextual answers. These responses combined text explanations with supporting charts and visualizations, which helped users grasp trends and patterns without having to interpret complex spreadsheets.

Making Complex Insights Easy to Understand

One of the main design challenges was balancing depth with simplicity. Financial data is inherently complex, but presenting it that way can intimidate users. To solve this, I focused on delivering insights in layers. The initial response from the assistant always summarized key findings in plain language. Users who wanted more detail could expand the result to view deeper analysis, specific metrics, or historical comparisons.

For example, instead of showing a dense table of numbers, the assistant might respond with something like:

“Your operating expenses increased by 15% compared to the previous quarter. Most of this increase came from higher marketing spend, which is up 22%.”

This approach made data easier to interpret and gave users the confidence to act on the information they were seeing.

Turning Complex Financial Insights Into Clear, Actionable Conversations

When I started working on this project for Midday, the main objective was to design an AI-powered assistant that made it easy for users to understand and interact with their financial data. Midday is a platform that provides businesses with deep financial insights, but many users struggled to turn that data into decisions. They needed a more intuitive way to explore their information without digging through multiple dashboards or manually interpreting charts. My goal was to design an experience that transformed complex data into a conversation that anyone could understand.

Designing Data Conversations Instead of Dashboards

The first major shift I focused on was changing the way users interacted with information. Instead of relying on traditional reports and static dashboards, I designed an experience where users could simply ask questions in plain language and receive meaningful answers. This meant building an interface that felt conversational and natural while still being powerful enough to deliver real insight.

Users could type questions like “What was my revenue growth last quarter?” or “How does my spending compare to last year?” and the AI would respond with clear, contextual answers. These responses combined text explanations with supporting charts and visualizations, which helped users grasp trends and patterns without having to interpret complex spreadsheets.

Making Complex Insights Easy to Understand

One of the main design challenges was balancing depth with simplicity. Financial data is inherently complex, but presenting it that way can intimidate users. To solve this, I focused on delivering insights in layers. The initial response from the assistant always summarized key findings in plain language. Users who wanted more detail could expand the result to view deeper analysis, specific metrics, or historical comparisons.

For example, instead of showing a dense table of numbers, the assistant might respond with something like:

“Your operating expenses increased by 15% compared to the previous quarter. Most of this increase came from higher marketing spend, which is up 22%.”

This approach made data easier to interpret and gave users the confidence to act on the information they were seeing.

a cell phone on a white block
a cell phone on a white block
a cell phone on a white block
two cell phones on a gray surface
two cell phones on a gray surface
two cell phones on a gray surface

Building Trust and Transparency Into the Experience

Trust was another key part of the design. Because users were making decisions based on AI-generated insights, they needed to know where the information was coming from and how it was calculated. I addressed this by including transparent design elements throughout the experience. Each insight included references to the underlying data source, links to supporting reports, and timestamps showing when the data was last updated.

The assistant also explained how it reached certain conclusions, especially for more complex queries. This helped build confidence in the system and positioned the AI as a collaborative tool rather than a mysterious black box.

Delivering Real Business Impact

The impact of this new experience was significant. After launch, users were able to get meaningful answers to their financial questions in seconds instead of minutes. Time-to-insight decreased by more than 60%, and engagement with advanced analytics features increased by over 3.5 times. Feedback from customers consistently mentioned how much easier it was to understand their financial position and make better decisions.

Most importantly, the AI assistant changed how people used Midday. Instead of logging in just to check reports, users began exploring their data more actively, asking follow-up questions, and uncovering insights they might have otherwise missed.

a cell phone leaning on a ledge
a cell phone leaning on a ledge
a cell phone leaning on a ledge

Outcome

This project transformed Midday from a static analytics tool into a dynamic and interactive financial intelligence platform. By focusing on simplicity, transparency, and conversational design, I created an experience that made complex data feel approachable and actionable. The result was a product that not only empowered users to understand their financial health but also helped them make smarter, faster business decisions.

Outcome

This project transformed Midday from a static analytics tool into a dynamic and interactive financial intelligence platform. By focusing on simplicity, transparency, and conversational design, I created an experience that made complex data feel approachable and actionable. The result was a product that not only empowered users to understand their financial health but also helped them make smarter, faster business decisions.

Copyright 2025 by Trey Underwood

Copyright 2025 by Trey Underwood

Copyright 2025 by Trey Underwood