Talkio Assessment Module Design
The Talkio Assessment Module ("Progress" tab) is a multi-dimensional AI-powered assessment module for Berlitz's new conversation practice tool, showcasing quantitative metrics as well as qualitative, personalized feedback. Through need analysis, SME consultation, and internal brainstorming sessions, I designed and developed a comprehensive 30+ page assessment guideline with 20+ user performance indicators and high-fidelity, interactive Figma prototypes to visualize user progress and competency.
Project type: Learning + Website Design
Role: Learning Experience Designer (Contract/Intern)
Industry: Language Learning, Adult Learning
Background
Berlitz is a renowned language learning platform with over 150 years of experience, specializing in immersive language instruction for organizations and professionals. Their unique pedagogy, the Berlitz Method, emphasizes immersive learning and learner-centric experiences in instructor-led courses and asynchronous training.
Recently, Berlitz has started developing their AI conversation tutor, a significant part of the personalized training offerings, to provide learners more flexibility to practice. I was tasked to redesign the website interface for a user-facing progress assessment tab. This project aims to enhance the user experience: provide more meaningful metrics and actionable feedback for learners to track their language proficiency progress and determine their next steps.
Delivery
Audit & Analysis
To understand the pain points with existing features and assessment models, I initiated a cross-functional consultation to consolidate issues from the development team and the client service side. The content design diagram on the right is used as a scaffold to help translate the existing assessment structure to a digestable brainstorming canvas, categorizing the existing content
and highlighting the revamps.
Assessment Mechanism Design
After understanding the precise needs, I dived into the assessment model design, from structuring the overall algorithm to defining sub-metrics and formulas.
First, based on Berlitz's current scoring system and CEFR standards, the I customized a Proficiency-Based Mechanism that caters to Berlitz's vision of offering personalized training: When assessingthe learner's performances, the choice of emphasized metrics should be based mostly on the learner’s Proficiency Level with the language practiced.
Next, I specified the 5 dimensions of conversation evaluation that aims to provide comprehensive feedback on learners' performances during practices, while highlighting their every growths and misses.
Finally, I created a 30+ page assessment guideline with 20+ user performance indicators to specify the definition and application of quantitative assessment with technical details and qualitative assessment with sample AI prompts for further development and testing.
Hi-Fidelity Figma Prototype
To help visualize the user journey and assessment system, I designed an interactive website prototype for the entire assessment ("Progress") tab. The site takes the format of a data-rich dashboard with customizable, expandable widgets that allow learners to get a comprehensive overview as well as dive into details.
I tailored each performance dimension widget with the corresponding indicators and next steps, while maintaining overall layout consistency.
