Quick Pulse: Live Classroom Feedback

Quick Pulse transforms classroom feedback into live, actionable insights. Teachers see exactly where their students stand, making lessons smarter, more engaging, and perfectly paced.

Comment 1

Inspiration

In our Calculus class, we noticed a persistent gap: teachers struggled to gauge real-time understanding, and students often silently fell behind. Traditional surveys were slow, often inaccurate, and failed to capture nuanced comprehension. Our teacher implemented a phone scanning system to anonymously gauge students' answers to multiple choice warm-up problems and he would write the stats for each answer on the board. A majority of students would not put up their QR code answer, because many of them did not want to openly admit they were thoroughly confused. We ourselves have experience this moment in between where we must decide to stay silent and be lost, or speak up and lose our intellectual dignity to our judging peers. We wanted to address this double-edged problem: students need help but often don’t voice it, and teachers lack actionable insights.

Our goal became clear: create a student-centered, teacher-empowering tool that provides real-time feedback, enabling dynamic, data-informed instruction.

From day one, we thought through edge cases and usability rigor:

  • Students can only answer a question once to prevent skewed analytics.
  • Unsaved sessions automatically go to drafts, kept aside until published to avoid confusion or lost data.
  • Initially, a teacher login separated access, but we temporarily disabled it to streamline early testing.
  • In production, we envision engagement analytics, trend detection, keyword aggregation, and adaptive suggestions for hyper-informed teaching.

- The answers can be cleared for multiple time usage on the same session, so if one round of survey shows low understanding, the teacher uses the feedback, and creates another survey with two clicks, the updated results are not skewed.

What it does

Quick Pulse is a live classroom feedback system that captures comprehension in real-time:

  • Students join via a class code.
  • Respond to a single comprehension prompt:
    • 🟢 “Good - solid understanding”
    • 🟡 “Iffy - eg. need more examples”
    • 🔴 “Lost - this requires more time”
  • Optional written feedback triggers only for “Lost” to minimize friction and unnecessary questions from students who understand.
  • Teachers receive instant analytics, including:
    • Response distribution percentages
    • Written comments
    • Session history and trend graphs

The real-time percentages are computed dynamically as:

$$ \text{Percentage of Good responses} = \frac{\text{# Good responses}}{\text{Total responses}} \times 100 $$

$$ \text{Percentage of Iffy responses} = \frac{\text{# Iffy responses}}{\text{Total responses}} \times 100 $$

$$ \text{Percentage of Lost responses} = \frac{\text{# Lost responses}}{\text{Total responses}} \times 100 $$

How we built it

Tech Stack

  • Languages: JavaScript, HTML, CSS
  • Frontend Framework: React
  • Backend / Database: Firebase Firestore
  • Authentication: Firebase Auth (email + session code login)
  • Design / Prototyping: Figma , Claude AI
  • APIs / Cloud Services: Firebase Realtime Listeners
  • Version Control / Collaboration: GitHub , Claude AI

Implementation Highlights

  • React state updates teacher dashboards instantly on submission.
  • Firestore listeners enable fully real-time updates across sessions.
  • Conditional rendering shows comment boxes only for 🔴 “Lost” responses.
  • Percentages and summaries are computed dynamically on the frontend for maximum responsiveness.
// Example: calculating Good responses percentage
const goodPercentage = (responses.filter(r => r.status === "Good").length / responses.length) * 100;

## Challenges we ran into
- **Real-time state management** initially caused dashboards to lag; integrating Firestore listeners solved this but required careful data normalization.  
- **Conditional comment logic** was deceptively tricky; ensuring the comment box triggers only for “Lost” without breaking submission flow required iterative debugging.  
- **Prompt caps and design instability**: Our initial Figma mockup was nearly functional but crashed repeatedly, consuming tokens and forcing redesigns.  
- **Code instability**: random frontend errors appeared during dynamic updates, necessitating precise handling of asynchronous state changes.  
- **Time constraints**: pivoting to a **teacher/student-centric problem** enhanced impact but reduced buffer for debugging and polish — a double-edged sword highlighting the tension between usability and relevance.  
- **Collaboration friction**: merging asynchronous GitHub changes required rigorous workflow coordination and strict code reviews.  
- **AI Tool Use Issues**: attempting to build a boilerplate with Claude was especially difficult as the prompt limit slapped us in the face numerous times. Completing our assigned tasks required a sufficient amount of account hopping and prompt optimization to achieve goals of convenience in website developing process. 

---

## Accomplishments that we're proud of
- Built a **fully functional, real-time classroom feedback system** under tight deadlines, especially with a later start.  
- Delivered a **polished, intuitive UI/UX**, integrating color-coded feedback, badges, and live trend analytics.  
- Implemented **robust safeguards** including single-answer enforcement, drafts, and conditional comment logic.  
- Established groundwork for **future advanced analytics**, including adaptive learning recommendations and engagement tracking.  
- Worked around loopholes such as stopping students from taking the same survey twice without permission, saving items in drafts, archiving completed surveys while maintaining statistics in archive, and adding a fully working log in system with approved codes and transfers of name data between the student and teacher accounts.
- Able to creatively develop methods of troubleshooting AI tools within limited time.

---

## What we learned
- Real-time feedback profoundly **enhances teacher insight** and allows lessons to adapt on the fly.  
- Handling **edge cases and conditional logic** is essential for a robust, production-ready system.  
- **UI/UX micro-optimizations**, such as color-coding, badges, and immediate analytics, substantially improve engagement.  
- **Collaboration discipline** and GitHub workflow rigor are critical in distributed development.  
- Thoughtful design choices, such as single-answer enforcement, drafts, and selective comment prompts, **bolster trust and data integrity**.  
- AI Prompts are limited without subscription, so choose wording and task compilation wisely. 
- API keys for certain companies require paid credits, so it is key to use open source models with unpaid API keys such as Gemini Flash 2.0 (example)

---

## What’s next for Quick Pulse
- Reinstate **teacher login and robust security** to prevent unauthorized access - removed for testing purposes.  
- Implement **keyword aggregation from comments** to accelerate problem detection.  
- Add visually appealing**trend charts and engagement analytics** to track understanding over time, and identify correlations between lesson and student understanding (eg. a teacher changes teaching methods and the student percentage for understanding graphically spikes)
- Explore **customizable questions, multi-step surveys, and adaptive recommendations** for differentiated instruction, including emoji sliders, multiple choice questions, long answer questions with AI integrated evaluation system, and more specified features for certain subjects.
- Incorporate **gamification and participation badges** to motivate students, encouraging active engagement and consistent feedback submission. Beyond badges, students could earn **achievement tokens** for consistent reflection, **streaks** for regular participation, or **level-ups** that unlock optional hints or extra resources. Mini **challenges** could be presented after a lesson for voluntary practice, and students could collect points that contribute to a **personalized progress tracker** rather than a public leaderboard, keeping the focus on self-improvement. These elements reward engagement without creating stress or competition, helping students feel recognized for their effort and building confidence in their understanding of the material.
- Scale Quick Pulse into a **full-featured, real-time teaching analytics platform**, helping educators teach smarter, faster, and more responsively.

Built With

Share this project:

Updates