The world of education can feel isolating and overwhelming for educators without mentors or role models.
We designed an AI-powered instructional Coach to enhance educators' coaching experience and address the challenges of traditional coaching methods.
Duration
30 Weeks
Team size
4 Members
Domain
Education
Role
UX Designer, Researcher
Tools
Figma, Miro, Figjam, Google Suite
CLIENT
We partnered with Logan & Friends, led by Dr. Jocelyn Logan Friend, who had a vision: to transform the coaching experience for educators using the power of AI.
VISION
VISION
Dr.Logan already has a vision of how this would be achieved:
Analyzing audio recordings of teaching sessions, comparing them against established teaching frameworks, and generating personalized feedback by combining AI capabilities with human expertise.
OBJECTIVE
OBJECTIVE
To build the experience around this vision for MVP1.
CHALLENGES
But translating this vision into a practical tool wasn't straightforward. Traditional coaching methods posed significant challenges.
01 Creating distress in the teaching environment
02 Limited observation time
03 Inadequate progress tracking
ADDRESSING THESE CHALLENGES
SOLUTION
"How might we redefine the coaching experience for educators?"
PROCESS
How did we approach them?
SOLUTION
O1
O1
Understanding the project scope
Weeks 1-3
Problem Identification
Problem Identification
Vision
O2
O1
Empathize with educators
Weeks 3-5
Problem Identification
Literature Review
Competitive Analysis
Define Interview
Participants
Conduct Interviews
Surveys
O3
O1
Define and Plan
product requirements
Weeks 6-8
Problem Identification
Affinity Maps / Data Analysis
Personas
User Journey Maps /
Use Cases
Project Requirements
O4
O1
Brainstorm and
Ideate
Weeks 9-14
Problem Identification
Storyboards
Mind Mapping
Develop Information Architecture
Create User Flows
Develop Wireframes
O5
O1
Design and
prototype
Weeks 15-27
Problem Identification
Low-Fidelity Prototypes
Design Validation / Iteration
Design System /
Style Guide
Mid/High-Fidelity Prototypes
O6
O1
Test
prototype
Weeks 28-30
Problem Identification
User Testing
Iterations
Handoff
EMPATHIZING WITH EDUCATORS
SOLUTION
We discovered that many teachers faced time-consuming processes and delayed feedback. Traditional coaching sessions often didn’t align with their busy schedules and provided generalized, sometimes subjective feedback. We knew we needed to address these issues with precision.
SOLUTION
SOLUTION
We focused on creating a user-friendly, intuitive, and supportive platform for teachers and coaches.
We addressed specific pain points and met their needs through various design and functional elements that empowered them to create engaging and effective learning experiences for their students.
Painpoint 1
Time-Consuming Process: Traditional coaching often involves scheduling meetings, which can be time-consuming and may not always align with teachers' busy schedules.
Painpoint 2
Limited Feedback Timeliness: Feedback from traditional coaching sessions may not be provided promptly, delaying teachers' ability to make immediate improvements.
Painpoint 3
Lack of Personalization: Coaching sessions may not always be tailored to individual teacher needs, leading to generalized feedback that does not effectively address specific challenges.
Painpoint 4
Subjectivity in Feedback: Feedback provided by human coaches may vary in quality and objectivity, depending on the coach's experience and perspective.
Painpoint 5
Limited Accessibility: Access to coaching sessions may be restricted by geographical location or availability of qualified coaches, limiting opportunities for professional growth.
Though our client had the vision in mind, the development process proved more complex than anticipated.
01 The team had no prior experience working with ML or AI models, so there were many questions about the feasibility, technical debt, and overall viability of the project.
02 Since there was no video input, It was challenging to understand how the AI model would use only audio to analyze classroom dynamics and provide accurate feedback to instructors.
To overcome this, We studied ML/AI concepts online, consulted university SMEs, and learned about Acoustic Feature Extraction, Voice/Sentiment Analysis, and NLP models.
DERIVING THE SOLUTION
Teachers would use an AI assistive tool to receive feedback.
Teachers were comfortable being recorded in class.
Educators believe that technology, particularly AI, should be an integral part of the classroom experience.
01 Empathy
Recognizing the challenges teachers face in balancing the classroom demands with their professional growth.
USER
RESEARCH
LITERATURE
REVIEW
CHALLENGE
PROBLEM
COMPETITIVE
ANALYSIS
POSSIBLE
SOLUTION
PROJECT
DEFINITION
Problem Statements
EMPATHIZING WITH EDUCATORS
MOTIVE
Gain foundational knowledge about instructional coaching, current practices, and the potential applications of AI in education.
GOAL
Understand the key challenges and opportunities within the field, identify best practices and pitfalls in designing AI-powered educational tools, and address technical and ethical challenges. This comprehensive understanding was essential to developing an effective and innovative AI Instructional Coach platform.
PROCESS
Conducted in-depth desk research, user interviews, surveys, and competitor analysis to gather deep insights into user needs, pain points, and behaviors to ensure the AI tool is user-centered and effectively addresses real problems.
Problem Statements
Schools contemplate using AI but haven't solidified plans yet.
Schools provide teachers with instructional coaches in America.
Teachers said they currently receive feedback verbally
KEY TAKEAWAYS
KEY TAKEAWAYS
01 AI can offer teachers enhanced support and feedback, fostering the development of effective teaching strategies.
01 Addressing teachers' intrinsic motivation and providing high-quality professional development resources are cruicial for their engagement and growth.
03 AI should not be seen as a replacement for human coaches, but rather as a tool to support teachers.
04 Core Challenges and Needs:
Teachers face significant challenges related to privacy concerns, technological adoption, timely feedback, and personalized professional development.
05 Pain Points and Barriers:
Traditional coaching methods are time-consuming, often provide delayed and generalized feedback, and can be costly and inaccessible for many teachers.
06 Opportunities - Technology Integration:
A significant percentage of educators are open to integrating AI and other technologies into their teaching practices, provided their concerns about privacy and effectiveness are addressed.
CREATING A USER-CENTRIC APPROACH
We created detailed personas, use cases, and scenarios that capture the varied needs, goals, and pain points of different educators and coaches. To keep the design process user-centered, we represented different users and archetypes using diverse data points from our research, updated the personas, and brainstormed collaboratively.
Personas, Use cases, Scenarios & Journey map
Challenge: Ensuring comprehensive use cases and scenarios was challenging. However, we overcame this by regularly validating use cases with stakeholders and real users.
02 Define
Translating research findings into actionable insights and potential solutions.
BREAKING DOWN THE PROBLEM STATEMENT
KEY TAKEAWAYS
How might we redefine the coaching experience for educators?
KEY TAKEAWAYS
How might we increase access to personalized instructional coaching for educators who face limitations due to budget, time, or lack of qualified coaches?
How might we create a feedback process for educators that is more frequent, actionable, and aligned with established teaching frameworks?
How might we foster a supportive environment where educators can easily access personalized resources and tools to continuously develop their skills and teaching strategies?
USER ROLES & PERMISSIONS
Given the product's complexity, we identified key user roles and tailored the design for the MVP's most essential roles to ensure the first release met the needs of our target audiences (Teachers and Coaches).
USE CASES & SCENARIOS
We defined specific situations and grouped them based on themes like awareness, consideration, adoption and onboarding, using the platform, implementing feedback, ongoing use and development.
Alex, 29
Elementary School teacher
A novice educator who is passionate and dedicated, skilled in elementary subjects, eager to learn, and ready to inspire students, yet unsure how to leverage technology effectively.
Ethan, 58
Instructional coach
With 25+ years of experience, mentors new teachers and excels in curriculum development and instructional coaching, fostering collaboration and lifelong learning to empower educators and enhance education.
Nicole Anderson, 45
Highschool teacher
An extroverted, experienced high school science teacher from Iowa, engages students with practical learning, 15+ years dedicated to dynamic education, continually updating her skills and tech integration.
Maria Gonzales, 37
Experienced Elementary School Teacher
A seasoned educator with 10 years, specializing in English, blending music, wellness checks, group activities, and assignments guided by curriculum objectives.
USER STORIES
We translated abstract user needs and project goals into concrete, actionable requirements to identify and prioritize the most important features and functionalities where users would interact with the AI tool to ensure it meets their needs and provide clear examples of how our proposed solutions would function in real-world situations.
1a
1b
INFORMATION ARCHITECTURE
Building the information architecture took some time. Now that we have identified all our product requirements and features. We needed to determine how to organize and structure our information to make it easy for users to navigate and find what they needed.
We encountered several obstacles. Some of the major issues were:
01. (1a) AI assistant and (1b) AI chat for feedback.
Determining whether these agents should be distinct entities or part of a unified system.
Pivot point: Considering the complexity of building the AI agent itself and the project duration we decided to move ahead as independent entities.
02. My lists and Calendar
The initial direction was to have a unified calendar that includes AI-generated suggestions or tasks for self-improvement and a section for instructor-written notes as well.
Pivot point: Since that was a lot to accommodate in the calendar, the team decided these should be made into a separate section called 'My Lists.'
At what point does an educator upload lesson plans?
Can the AI re-analyze the report after the lesson plan has been uploaded at a later point?
Decision making: Does user need the option to input class details before starting recording and after recording
Can AI suggest grade and class titles based on the teacher's calendar?
Record flow
Data input
Record
Report generation
Upload recording
PROCESS FLOWS
We then Mapped out the step-by-step sequence of actions users take to complete tasks and identified bottlenecks or pain points, detecting potential areas where users might experience confusion or frustration.
Decision: After careful consideration, we scoped down the project requirements to accommodate the new developments and eliminated my list, coaching & performance for MVP1.
PARALLEL DECISION MAKING POINTS
01 Teaching Framework Integration
Since our project scope doesn't involve developing or training AI models or working with teams that do. Our only way forward was to build the experience based on the assumption of AI-powered feedback based on teaching frameworks that L&F might accommodate in the future.
02 AI suggested resources
The client was bringing in a third-party team to develop content and learning resources and build the entire resource hub for the platform.
ITERATING THE FEEDBACK PAGE
Reports page
Feedback page
Overview page
Schedule session modal
WIREFRAMES
We then established the basic layout and structure of each screen, ensuring that all necessary elements are included and logically arranged.
Personas, Use cases, Scenarios & Journey map
03 Ideate
Brainstorm and sketch out ideas to visualize the user interface and experience.
1
2
3
I divided the feedback page to accommodate:
'Audio Recording' view to enable educators to revisit the classroom session.
Section to see AI-generated feedback at different time stamps.
Each feedback has a chat space to engage in a dialogue with AI coach.
1
Issue: Considering the amount of feedback the AI coach might give and educators engaging in chat with the AI coach.
Increases scroll time which makes it overwhelming for teachers.
Makes it hard to identify the most important feedback.
Could distract from other feedback channels due to written comments.
Not effective in terms of finding time to engage with the AI coach.
Selected
To solve this issue I reiterated the feedback page to accommodate a dynamic chat feature that changes chat content based on the selected feedback.
To further ease the scrolling experience and identify critical feedback, I implemented time stamps on the audio timeline so teachers can quickly navigate to specific feedback.
CLASS SESSION RECORDING
Teachers can record their classroom sessions directly through the platform. These recordings are then stored and organized in a dedicated reports page.
Personas, Use cases, Scenarios & Journey map
DETAILED, TIMESTAMPED FEEDBACK
Timestamped feedback directly linked to the audio recording, allowing teachers to listen to the exact instances highlighted in the input. Additionally, the platform includes an interactive chat feature with AI, enabling teachers to discuss the feedback in real time.
INTEGRRATED LEARNING HUB
The hub offers tailored resources, such as articles, videos, and online courses, specifically chosen to address the feedback areas. Teachers can use these resources to develop skills in targeted areas, further enhancing their teaching practices.
REPORT MANAGEMENT SYSTEM
Recordings are stored and organized in a dedicated reports page, which features robust filtering options to help teachers navigate through their past evaluations. This system simplifies the review process and helps teachers access their performance data over time efficiently.
TEACHER'S DASHBOARD
The product provides an overview of the teacher's performance for a particular duration, number of sessions, number of flagged reports. Performance across all domains from the teaching framework, calendar, activities and upcoming sessions
04 Design
Create visual and interactive representations of the tool by developing wireframes, prototypes, and visual designs.
FROM CONCEPT TO PROTOTYPE View Prototype ↗️
Teachers can quickly navigate by clicking on the time stamp, view feedback, and engage in chat with AI.
Other considerations:
1. Multi-modal chat experience to recommend learning resources.
2. Tying feedback to teaching frameworks to signify improvement in specific categories.
(Color coordination, Tags)
E. Performance page
Issue: The performance page lacks clarity regarding user scores.
Percentage indications are missing for each card.
Color-coding to represent different domains is absent.
Understanding subdomain scores is difficult.
Recommendation: Display percentages on each performance card.
Use color-coding to visually represent different domains.
Improve the presentation of sub-domain scores for better comprehension.
USABILITY TESTING
For the usability testing, we created a detailed user study plan. We conducted various tests, including expert review, think-aloud sessions with experienced teachers, novice teachers, and teacher coaches, internal evaluation through cognitive walkthroughs and design critique, and finally, a user satisfaction questionnaire.
05 Testing
Scroll
Issue: The current scroll and fixed layout on the feedback page reduces the scrollable area, hindering user experience.
Recommendation: Re-evaluate the layout to maximize the scrollable area for easier reading of feedback.
B. Flagged feedback
Issue: Users cannot easily identify flagged feedback items due to a lack of visual cues.
Recommendation: Implement a review status indicator on each feedback card to highlight flagged items.
C. 'Learn more' option
Issue: The "Learn More" option followed by hover functionality for accessing feedback details seems redundant.
Recommendation: Remove the "Learn More" text. Users should be able to click directly on the feedback card to access details.
D. AI assistant icon
Issue: The AI Assistant icon looks like a comment bubble, potentially leading to confusion with the feedback section.
Recommendation: Redesign the AI Assistant icon to clearly indicate it as the functionality to chat with the AI bot.
My Learnings
The project's non-linear progression was a natural consequence of several factors. Our team's initial inexperience with AI and ML models presented unforeseen technical challenges, especially with audio analysis and AI feedback mechanisms. This required revisiting earlier design stages to refine our approach and adapt to the technology's complexities.
Simultaneously, the project scope evolved as we gained deeper insights into the technology and user needs through extensive research and interviews. For instance, the decision to separate the AI assistant and chat functionality, as well as refinements to the information architecture, stemmed directly from these learnings.
Our commitment to an iterative design process, incorporating user feedback through prototyping and usability testing, further emphasized the non-linear nature of the project. We embraced this flexibility, recognizing that continuous refinement based on real-world data would lead to a more effective and valuable product.
Additionally, adopting an agile methodology allowed us to prioritize the MVP and adapt to changing requirements throughout development. This approach, combined with our research-driven design philosophy, led to a fluid design process where initial concepts were continually revisited and refined based on new insights and evolving needs.
I'm glad you made it here; let's connect.
