Timeline
October 2025 - December 2025
Role
UX/UI Designer
Tools
Figma, Miro, Canva
PROBLEM
ACL Calculation Experts struggle to streamline decisions
ACL calculations are required for every product changeover (thousands per year) and must be documented for the lifetime of the product. However, calculations are done manually, one by one using an Excel sheet to input all the data. Currently, the R&D Scientist is the one primary expert who performs nearly all ACL calculations.
These calculations are very time-consuming and require technical understanding of ecotoxicology, product composition, manufacturing context, and analytical detection which causes delays in approvals and makes it prone to errors.
SOLUTION
AI-Assisted Calculation
Created an intuitive AI-assisted workflow minimizing the user's steps. Rather than requiring users to manually validate calculations, we designed a system to use three AI agents that communicate with one another, reducing cognitive load and calculation errors.
AI System:
Calculation Agent: Runs the ACL formula
ACL Expert Agent: Reviews outputs using domain-specific rules
Lead Agent: Coordinates validation and delivers a clear, user-facing result
Review:
Check for errors within the AI's calculation
Ensures the user if the product has toxicity using color (green/red)
Gives final results on the measuring requirements for the products.
RESEARCH
User Needs and Personas
Conducted an interview with the Research and Development Scientist and she offered insightful information on the company's existing ACL workflow. We were able to understand her pain points and desired improvements.
While researching, we found that currently, there are only a few backups who are familiar with the process but mainly done by a single specialist, the R&D Scientist. A calculation can be finished in about ten minutes when all required data is in the database, however, it can take up to two days when new data needs to be sourced. The total burden is fundamental because thousands of product changes require ACL calculations.
Through our research, we created three personas that captured the individuals that interact with ACL calculation.
IDEATION
AI Integration
During the design phase, we explored different ways to integrate AI into the system to support accuracy and efficiency.
AI Agents
One early concept used two AI agents: one to run the ACL calculation and a second to independently verify the result. This approach directly addressed the need for error-proofing in a calculation heavy workflow.
ChatBot
Include a chatbot for immediate help with formula understanding.
Review
Users review the AI's calculation before submitting for approval.
Design
Color scheme:
High Fidelity Wireframes:
REFLECTION
My Learning Output
This was my first time prototyping using Figma, so getting the hands-on experience was a huge learning curve and made me fall in love with Figma as an application.
I was in charge of the interactions and design for our prototype and although it took me two nights, it was all worth it when we presented our final to the R&D Scientist and our professor. This project is one of the reasons I became passionate about UX design.