Case Studies
Apr 18, 2025

Robert Grashuis
OneSpring Partner & CXO
Thank My Teacher Initiative
In an era where digital communication often takes precedence over face-to-face interactions, the team behind "Thank My Teacher" recognized a significant need in the educational community.
Our mission was to create a specialized platform that not only utilizes state-of-the-art technology but also retains the emotional warmth and sincerity often missing in digital exchanges.
This initiative was particularly geared towards enabling students, parents, and fellow educators to express their heartfelt gratitude to teachers. By merging modern tech with the universal and timeless act of saying "thank you," we aim to elevate the experience of giving and receiving thanks within the education sector.
Challenge
The imperative for the Thank My Teacher team was clear:
Provide assistance when composing the "thank you" message to the teacher.
Leverage artificial intelligence to identify and flag inappropriate, negative, or vulgar language within the thank-you notes.
These goals presented a unique set of challenges, such as elevating the quality of digital communication, to ensure users can convey their gratitude effectively and genuinely.
Also, the complexity associated with handling intentionally misspelled profanity or symbols replacing standard characters—a common occurrence in contemporary online communications.

Figure 1.0 - Thank My Teacher home page
Solution: "Thank My Teacher" Writing Assistant
Two-Fold Strategy
Our solution approach was two-fold. First, provide a "Thank My Teacher" writing assistant to help the student or parent create a meaningful and genuine message of gratitude. Our approach was to create a writing assistant that integrates AI to elevate the quality of digital communication, ensuring that users can convey their gratitude effectively and genuinely.
Key Features
Sentiment Analysis
A standout feature is its sentiment analysis capability. As a user drafts a thank-you note, the AI can promptly identify and highlight any negative undertones, advising on potential revisions to better align the message with its intended sentiment of appreciation. This isn't just about censoring or correcting; it's about helping users effectively communicate their feelings and achieve the desired impact.
Personalization Options
What adds depth to this tool is the user's ability to tailor their message according to specific sentiments. They can emphasize aspects like:
Support
Expertise
Dedication
Patience
Instances of going beyond the norm
Such granularity ensures that users aren't just penning generic thank-you notes, but ones that genuinely reflect their experiences.

Figure 2.0 - Thank My Teacher Writing Assistant
Customization Features
The Assistant offers customization in terms of the note's length, catering to those who prefer succinct acknowledgments as well as those inclined to provide a detailed account. Through these features, the "Thank My Teacher Writing Assistant" doesn't merely automate gratitude but enriches the entire process of articulating appreciation.
Through specifically crafted API endpoints, the AI evolves from a mere tool to a caring companion. It aids, assists, and sometimes even co-writes, ensuring each message is nothing short of heartfelt.
AI-Driven Sentiment Analysis
Second, create a sentiment analysis engine driven through AI language evaluation to identify and guide corrective action. Our approach was to apply sentiment analysis, capitalizing on the capabilities of AI's advanced language models to address these challenges.
Processing Workflow
Upon a user submitting a thank-you note, the text is initially directed to the sentiment analysis API. This generates crucial data points, such as tone, sentiment score, flagged profanity, and potential suggestions for improvement.
AWS Lambda functions were employed to further tailor the AI prompt, facilitating communication with OpenAI's ChatGPT model. The ultimate decision to flag a note as inappropriate relies on a composite evaluation of both sentiment score and profanity indicators, all processed and displayed seamlessly on the frontend.

Results
A Fusion of Technology and Sentiment
Within the rapidly evolving sphere of educational technology, "Thank My Teacher" sets itself apart, serving as both a paragon of technical ingenuity and a channel for authentic human sentiment.
Development Journey
Our journey began with initial design blueprints created in Figma, then moved on to sophisticated back-end implementations utilizing WordPress and PHP.
Technical Innovation
Notably, a cornerstone of our platform's capabilities lies in the seamless integration of AI-driven API endpoints. These custom-built API endpoints serve multiple pivotal roles, from sentiment analysis to language enhancement, thus elevating the user experience to new heights.
Thanks to OneSpring's human-centered, research-driven approach, Stride now had:
AI-powered writing assistant with full API integration
Sentiment analysis engine driven through AI language evaluation
508c compliance for complete accessibility
A digital hub for gratitude in education