
Weekend Creative Challenge: HackSprint AI
HackSprint AI is a simple tool that helps students create hackathon project ideas quickly. Users enter their domain, skills, hardware, budget, and team size. The app generates three project ideas with a problem, solution, features, and innovation score. It uses AWS Lambda, AWS Amplify, and Amazon Bedrock. It makes finding and planning hackathon ideas easier and faster.
Weekend Creative Challenge: HackSprint AI - The Hackathon Project Idea Generator
Vision & What the App Does
HackSprint AI is a creative web application that instantly generates unique, tailored hackathon project ideas based on user preferences. As a student who participates in hackathons regularly, I noticed that teams waste 2-3 hours during the brainstorming phase — precious time that could be spent building. HackSprint AI eliminates this problem entirely.
How it works: Users fill a simple form with 5 inputs — Domain, Skills, Hardware, Budget, and Team Size. The app then generates 3 complete project proposals, each containing:
- 🎯 Project Title
- ❓ Problem Statement
- 💡 Proposed Solution
- ✅ Key Features (4 features per idea)
- 🛠️ Recommended Technologies
- 🚀 Innovation Factor
- 📊 Difficulty Level (Easy/Medium/Hard)
- ⭐ Scoring System (Innovation, Feasibility, Cost-Effectiveness, AI Potential — each scored 1-5, plus an overall score out of 10)
The app covers 9 hackathon domains: Healthcare, Education, Agriculture, Finance, Environment, Smart City, Security, Transportation, and Entertainment.
Expected Output
Input:
Field | Value |
|---|---|
Domain | Healthcare |
Skills | Python, AI |
Hardware | None |
Budget | Very Low (Free) |
Team Size | 4 |
Output:
The app generates 3 complete project idea cards — AI Health Symptom Checker, Smart Medicine Reminder, and Mental Health Mood Tracker — each with full details and scores.

-
- IDEA - 2

-
- IDEA - 3

-
The app produces different results for each domain. Selecting "Education" generates ideas like "AI Study Plan Generator," "Quiz Generator from Notes," and "Virtual Lab Assistant" — each tailored to education-specific challenges.
Development Process
I built HackSprint AI entirely from the Windows Command Prompt using the AWS CLI — no console clicking, no GUI tools. The entire project was built and deployed in under 3 hours.
Phase 1: Backend Setup
- Created an IAM execution role (
hacksprint-lambda-role) with Bedrock and Lambda permissions
- Wrote a Python 3.12 Lambda function that processes user inputs and returns intelligent, domain-specific project recommendations
- Deployed using
aws lambda create-functionwith a zip file
- Created a public Function URL with CORS support for frontend communication
Phase 2: Frontend Development
- Built a responsive single-page application with vanilla HTML, CSS, and JavaScript
- Designed a dark gradient UI with gold accents for a modern hackathon feel
- Implemented loading animations and dynamically rendered idea cards
- Connected frontend to Lambda via the Function URL
Phase 3: Deployment
- Hosted the frontend on AWS Amplify using CLI commands (
create-app,create-branch,create-deployment,start-deployment)
- Pushed all source code to GitHub for version control
Key Decisions
- Serverless-first: Lambda eliminates server management and costs $0 for low traffic
- Lambda Function URLs over API Gateway: Simpler setup, built-in CORS, one CLI command
- No frameworks: Vanilla JS keeps deployment simple and bundle size at zero
- Domain-specific intelligence: Built a comprehensive knowledge base for 9 domains rather than generic random outputs
Challenges & How I Overcame Them
Challenge 1: Bedrock billing requirement — My AWS account needed a payment instrument for model access. I solved this by designing the architecture with Bedrock integration code ready, while using a curated intelligent response system as the immediate data source. This taught me about building systems with graceful fallbacks.
Challenge 2: Windows CMD compatibility — JSON escaping and file operations behave differently on Windows. I used PowerShell commands (
Compress-Archive , Set-Content ) to handle file operations that CMD couldn't.Challenge 3: Lambda code deployment — Initially uploaded a nearly-empty file (91 bytes!). Caught the error by checking file sizes with
dir , then used update-function-code to fix it without recreating the function.
AWS Services Used / Architecture Overview
Architecture Diagram
┌─────────────────┐ ┌──────────────────┐
│ User Browser │────────▶│ AWS Amplify │
│ (Form Input) │ │ (Static Hosting) │
└─────────────────┘ └──────────────────┘
│
▼
┌──────────────────┐
│ Lambda Function │
│ URL │
│ (HTTPS + CORS) │
└──────────────────┘
│
▼
┌──────────────────┐
│ AWS Lambda │
│ (Python 3.12) │
│ Idea Generation │
└──────────────────┘
│
▼
┌──────────────────┐
│ Amazon Bedrock │
│ (Architecture │
│ Ready - Claude/ │
│ Nova Models) │
└──────────────────┘AWS Services Used
Service | Purpose | How It's Used |
|---|---|---|
AWS Lambda | Serverless compute | Runs Python 3.12 function that processes inputs and generates project ideas |
Lambda Function URL | API endpoint | Provides HTTPS endpoint with CORS — no API Gateway needed |
AWS Amplify | Frontend hosting | Serves HTML/CSS/JS globally with CDN |
AWS IAM | Security | Execution role with Bedrock + Lambda permissions |
Amazon Bedrock | AI integration | Architecture designed for Claude 3 Haiku / Nova Lite model invocation |
Technical Specifications
- Lambda Memory: 256 MB | Timeout: 30s | Runtime: Python 3.12
- Region: us-east-1
- Authentication: Public (for demo)
- Frontend: Vanilla HTML5 + CSS3 + JavaScript (ES5 compatible)
What I Learned
- AWS CLI Mastery: Building an entire application from the command line — IAM roles, Lambda functions, Amplify deployments — taught me how AWS services connect at a fundamental level.
- Serverless Cost Efficiency: The entire app runs within AWS Free Tier. Lambda's 1M free requests/month + Amplify's free tier = $0/month. Serverless isn't just easier — it's cheaper.
- Function URLs are Underrated: Lambda Function URLs eliminated the need for API Gateway entirely. One command gives you a production HTTPS endpoint with CORS.
- Fallback Architecture: When Bedrock access wasn't available due to billing, I designed the system with graceful degradation. The code supports AI model swapping — change one model ID and the entire pipeline works.
- Debugging in Production: Checking file sizes, using
update-function-codefor hot fixes, and understanding Lambda states taught me real-world debugging skills.
- End-to-End Ownership: Going from zero → working app → deployed → live URL → GitHub in one session gave me confidence to ship fast.
Links
1.🌐 Live App - https://main.d8d16mc4rdxks.amplifyapp.com
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article