DEV Community: Basel Mohamed Alam

🎓 Building an AI-Powered Study Buddy with AWS (Bedrock + Lambda)

Basel Mohamed Alam — Mon, 12 Jan 2026 16:23:17 +0000

Transform Your Study Materials into Interactive Quizzes with Claude AI

Being working on this App for a While,
Have you ever wished you could instantly generate quiz questions from your study materials? In this tutorial, I'll show you how to build a fully functional AI-powered study assistant using AWS services. This project is perfect for learning serverless architecture and working with generative AI!

🚀 What We're Building
Study Buddy is an AI-powered application that:

✅ Accepts study materials (PDFs, images, text files)
✅ Extracts text using OCR (Amazon Textract)
✅ Generates customized quiz questions using Claude AI (Amazon Bedrock)
✅ Allows users to specify difficulty level and question count
✅ Provides explanations for correct answers

Repository
https://github.com/basel5001/Study-buddy

🏗️ Architecture Overview

📌 Project Overview

Study Buddy is an AI-powered application that transforms study materials into interactive quiz questions. Students can upload documents, extract text, and get customized quizzes generated by Claude AI.

🎯 Key Features

Document Upload - Accept PDFs, images, and text files
Smart Text Extraction - OCR capability using Amazon Textract
AI Quiz Generation - Claude AI creates intelligent questions
Customizable Settings - Choose difficulty (easy/medium/hard) and number of questions
Interactive Quiz Interface - Answer questions and get instant feedback
Detailed Explanations - Learn why answers are correct
Score Tracking - See your performance immediately

🏗️ System Architecture

Components:

Frontend (React)
- User interface for uploading materials
- Quiz taking interface
- Results dashboard
API Layer (API Gateway)
- /upload - Handle document uploads
- /extract - Process text extraction
- /generate-quiz - Generate quiz questions
Processing Layer (AWS Lambda)
- Function 1: Extract text from documents
- Function 2: Generate quiz using AI
AI Services
- Amazon Textract: OCR and text extraction
- Amazon Bedrock (Claude): Intelligent quiz generation
Storage (S3)
- Store uploaded documents
Monitoring (CloudWatch)
- Track logs, errors, and performance metrics

🔄 User Flow
Step 1: Upload Study Material

User uploads a document (PDF/image) or pastes text
File is stored in S3 bucket

Step 2: Extract Text

Lambda function retrieves document from S3
Amazon Textract extracts all text content
Extracted text is displayed to user for review

Step 3: Generate Quiz

User selects preferences:
- Number of questions (1-20)
- Difficulty level (easy/medium/hard)
- Question type (multiple choice)
Text is sent to second Lambda function
Claude AI generates customized quiz questions with:
- Question text
- 4 multiple choice options
- Correct answer
- Explanation

Step 4: Take Quiz

User answers questions interactively
Selections are tracked in real-time
Submit when ready

Step 5: View Results

See correct/incorrect answers highlighted
Read explanations for each question
View final score and percentage
Option to generate new quiz

💰 Cost Breakdown (Monthly Estimates)
Light Usage (10 quizzes/day):

Bedrock (Claude): ~$3-5/month
Textract: ~$1-2/month
Lambda: Free tier (negligible)
S3 Storage: ~$0.50/month
API Gateway: ~$1/month
Total: ~$6-10/month

Medium Usage (100 quizzes/day):

Bedrock: ~$30-50/month
Textract: ~$10-15/month
Lambda: ~$5/month
Other services: ~$5/month
Total: ~$50-75/month

🛠️ AWS Services Used

Service	Purpose	Why This Service
S3	Document storage	Cheap, reliable, easy Lambda integration
Lambda	Serverless compute	No server management, pay per use
API Gateway	REST API endpoints	Managed API with built-in security
Bedrock	AI text generation	Access to Claude without managing infrastructure
Textract	OCR/text extraction	Purpose-built for document processing
CloudWatch	Logging & monitoring	Native AWS monitoring solution

🎨 Design Decisions
Why Serverless?

No server maintenance
Auto-scaling
Pay only for what you use
Perfect for sporadic usage patterns

Why Claude (Bedrock)?

Excellent at following instructions
Consistent JSON output
Strong reasoning for educational content
No need to manage API keys separately

Why React Frontend?

Component-based architecture
Easy state management
Fast development
Great user experience

📊 Technical Highlights
Lambda Function 1: Text Extraction

Trigger: API Gateway POST request
Process: Fetch document → Textract → Extract lines
Output: Clean text ready for processing
Timeout: 30 seconds
Memory: 512 MB

Lambda Function 2: Quiz Generation

Trigger: API Gateway POST request
Process: Format prompt → Call Bedrock → Parse JSON
Output: Structured quiz questions array
Timeout: 60 seconds
Memory: 1024 MB

API Gateway Configuration

Type: REST API
Authentication: None (can add API keys)
CORS: Enabled for browser access
Rate Limiting: Optional (10 req/sec recommended)
Stage: Production

🚀 Deployment Workflow
Setup Phase:

Enable Bedrock model access
Create S3 bucket
Create IAM roles with proper permissions
Deploy Lambda functions
Configure API Gateway
Deploy React frontend

Testing Phase:

Test each Lambda function individually
Test API endpoints with Postman/cURL
Verify CORS configuration
Test complete user flow
Monitor CloudWatch logs

Production Phase:

Add custom domain (optional)
Enable CloudWatch alarms
Set up usage plans and API keys
Configure backup strategies
Monitor costs and usage

🎓 Learning Outcomes
Skills Gained:

✅ Serverless architecture patterns
✅ AWS service integration
✅ AI/ML API integration
✅ Prompt engineering for LLMs
✅ RESTful API design
✅ Error handling in distributed systems
✅ CORS configuration
✅ React state management

Best Practices Applied:

Separation of concerns (frontend/backend)
Stateless Lambda functions
Proper error handling and logging
Environment variable usage
Security through IAM roles
Cost optimization strategies

🔮 Future Enhancements
Phase 1: Core Improvements

Add user authentication (AWS Cognito)
Store quiz history in DynamoDB
Support more file formats (DOCX, PPT)
Add true/false and fill-in-blank questions

Phase 2: Advanced Features

Spaced repetition algorithm
Study session tracking
Progress analytics dashboard
Collaborative study groups
Flashcard generation

Phase 3: Scale & Optimize

Mobile app (React Native)
Offline mode capability
Export to PDF/Anki
AI tutor chat interface
Voice mode with Amazon Polly

🐛 Common Challenges & Solutions
Challenge 1: CORS Errors

Problem: Browser blocks API requests
Solution: Enable CORS in API Gateway + return headers in Lambda

Challenge 2: JSON Parsing from AI

Problem: Claude wraps JSON in markdown
Solution: Use regex to extract pure JSON before parsing

Challenge 3: Token Limits

Problem: Large documents exceed limits
Solution: Truncate text to 4000 characters or chunk processing

Challenge 4: Timeout Errors

Problem: Lambda times out during processing
Solution: Increase timeout to 60s and optimize code

Challenge 5: Bedrock Access

Problem: "Access Denied" errors
Solution: Request model access in Bedrock console first

📈 Success Metrics
Technical Metrics:

Average response time: < 5 seconds
Error rate: < 1%
Quiz generation success: > 95%
Text extraction accuracy: > 90%

User Metrics:

Quiz completion rate
Average score improvement over time
Daily active users
Quiz variety (different subjects)

🎯 Key Takeaways
What Worked Well:

Serverless architecture kept costs low
Claude AI produced high-quality questions consistently
React provided smooth user experience
AWS services integrated seamlessly

What Could Be Improved:

Add caching for repeated requests
Implement batch processing for multiple documents
Add more question types
Better mobile responsiveness

Business Potential:

SaaS model for students/teachers
B2B sales to educational institutions
Premium features (advanced analytics, unlimited quizzes)
White-label solution for EdTech companies

📚 Resources & Next Steps
Documentation:

AWS Lambda Best Practices
Bedrock API Reference
React State Management
Prompt Engineering Guide

Next Steps:

Deploy a demo version
Gather user feedback
Iterate on features
Write comprehensive tests
Create video tutorial
Launch on Product Hunt

💡 Pro Tips

Start with the free tier to minimize costs
Use CloudWatch extensively for debugging
Test each component independently first
Keep prompts simple and clear
Always handle errors gracefully
Document your API endpoints
Version your Lambda functions
Set up alarms for critical metrics

🌟 Project Impact
For Students:

Faster exam preparation
Better retention through active recall
Personalized learning experience
Study anywhere, anytime

For Developers:

Real-world serverless experience
AI integration portfolio piece
Cloud architecture knowledge
Full-stack development practice

Total Build Time: ~8-12 hours
Difficulty Level: Intermediate
Tech Stack Complexity: Medium
Scalability: High
Maintenance: Low

AWS and Google Cloud officially announced a jointly engineered multicloud

Basel Mohamed Alam — Fri, 05 Dec 2025 00:39:35 +0000

Breaking News!!!

1️⃣ AWS and Google Cloud officially announced a jointly engineered multi-cloud networking solution, introducing a new open specification for network interoperability.

2️⃣ This collaboration brings together AWS Interconnect – multi-cloud and Google Cloud’s Cross-Cloud Interconnect, enabling customers to establish private, high-bandwidth, low-latency connectivity between both cloud providers with unprecedented speed, automation, and simplicity.

🌐Region Pairs Now in Preview

Support begins with the following AWS ↔️ Google Cloud region mappings:

US Regions

• AWS us-east-1 → GCP us-east4 (N. Virginia)

• AWS us-west-1 → GCP us-west2 (N. California ↔ LA)

• AWS us-west-2 → GCP us-west1 (Oregon ↔ Oregon)

Europe Regions

• AWS eu-west-2 → GCP europe-west2 (London)

• AWS eu-central-1 → GCP europe-west3 (Frankfurt)

🟢 This is a major shift in how enterprises can build true multi-cloud architectures, without the traditional complexity of physical circuits, manual routing, or long provisioning cycles. Connectivity can now be activated in minutes, fully managed by both cloud providers end-to-end.

reinvent #multicloud #AWS #GoogleCloudCloudComputing #GCP #CloudNetworking

Be Aware!! any one using npm pkgs

Basel Mohamed Alam — Sun, 30 Nov 2025 13:49:56 +0000

CloudDefense.AI

Sep 19 '25

Shai-Hulud: a self-propagating npm worm hits @ctrl/tinycolor and dozens more packages

#news #opensource #npm #security

Comments

2 min read

[Boost]

Basel Mohamed Alam — Mon, 13 Jan 2025 19:32:53 +0000

10 Ways To Secure Docker Images: Best Strategies

Fave😌✨ ・ Jan 6

#docker #security #devops #cloud

A Comparative Analysis of Terraform and CloudFormation

Basel Mohamed Alam — Mon, 06 Jan 2025 00:11:59 +0000

The way we deliver and manage infrastructure has been completely transformed by Infrastructure as Code (IaC). AWS CloudFormation and Terraform from HashiCorp are two of the best tools in this field. Let's examine their benefits, drawbacks, and distinctions as well as some real-world applications for each.

Terraform

Advantages:

Support for several Clouds: Terraform is cloud-agnostic, meaning you can use a single tool to manage infrastructure across several cloud providers (AWS, Azure, and GCP).
Rich Ecosystem: An extensive network of reusable modules and providers that improve efficiency and flexibility.
State Management: Keeps your infrastructure in good condition so you can easily handle complicated deployments. Cons: 1. Learning Curve: Novices may find the HashiCorp Configuration Language (HCL) difficult to understand.
Complexity of State Management: Managing state files, particularly Example Use Case: A company operating in a multi-cloud environment (AWS, Azure) can use Terraform to ensure consistent infrastructure provisioning and management across both platforms, simplifying operations and reducing the learning curve for their DevOps team. CloudFormation Pros:
Deep AWS Integration: Seamlessly integrates with all AWS services, leveraging the latest AWS features and best practices.
Free to Use:There are no additional costs for using CloudFormation; you only pay for the resources you provision.
Drift Detection: Detects changes made outside CloudFormation, helping maintain the integrity of your stacks. Cons: 1.AWS-Specific: Limited to AWS, making it unsuitable for multi-cloud strategies. 2.Verbose Templates: JSON or YAML templates can become cumbersome and difficult to manage for large deployments. Example Use Case: An AWS-centric organization can utilize CloudFormation to automate the provisioning of its infrastructure. For instance, deploying a complex web application stack (EC2, RDS, S3) can be efficiently managed and maintained using CloudFormation, ensuring tight integration and optimization with AWS services. Key Differences:
Cloud Support: Terraform is multi-cloud, whereas CloudFormation is AWS-specific.
Language: Terraform uses HCL, while CloudFormation uses JSON/YAML.
State Management: Terraform manages its own state files; CloudFormation relies on AWS to maintain stack state.
Community and Modules: Terraform has a broader ecosystem of community modules compared to CloudFormation. Choosing between Terraform and CloudFormation depends on your specific needs and environment. For multi-cloud flexibility, Terraform is ideal. For AWS-centric operations with deep service integration, CloudFormation is the way to go.

Power of Online Communities for Tech Learning

Basel Mohamed Alam — Sun, 05 Jan 2025 23:46:49 +0000

We discussed the benefits of online community participation for continuous learning and development in the tech field. You highlighted your experience with several platforms:
AWS Re:Post: Great source for AWS knowledge with insightful content and discussions.
DEV Community: Offers a welcoming space for developers to share, collaborate, and learn.
Stack Overflow: Unmatched expertise for troubleshooting and finding coding solutions.
Reddit: Specific subreddits provide diverse perspectives on tech trends, careers, and even humor.
Mastodon: A decentralized social network is gaining traction and offering a different approach to online discussions.
These communities offer several advantages for tech professionals:
Exposure to New Ideas: Encountering fresh perspectives and innovative solutions from others.
Problem-solving Prowess: Helping others on forums hones your troubleshooting skills.
Staying on Top of Trends: Accessing industry updates and insights keeps you ahead of the curve.
Networking Opportunities: Connecting with like-minded individuals can lead to collaborations and career growth.
The post concludes with a call to action, inviting others to share their favorite tech learning communities. Relevant hashtags enhance discoverability within the tech community on LinkedIn.

Feel free to Connect with me through my account:
https://lnkd.in/dd5KA-6K

Revolutionizing DevOps with AI: The Future is Here

Basel Mohamed Alam — Sun, 05 Jan 2025 23:39:28 +0000

In the fast-paced world of DevOps, where efficiency and reliability are everything, AI is proving to be a game-changer. From automating repetitive tasks to predicting potential bottlenecks before they occur, AI is helping DevOps teams focus more on innovation and less on firefighting.
This article by ClickUp How to Use AI in DevOps is a great resource to explore the possibilities AI brings to:
🔧 Continuous Integration and Deployment (CI/CD): Smarter pipelines for faster and error-free releases.
📊 Performance Monitoring: Predictive analytics to prevent downtimes and ensure smoother user experiences.
🤖 Automation: Let AI handle the mundane so we can focus on the strategic.
💡 Incident Management: Faster root cause analysis and remediation suggestions.
As a DevOps engineer, I'm always excited to explore how emerging technologies like AI can make our workflows smarter and more efficient. Let's leverage these tools not just to improve processes, but to reimagine what's possible in the DevOps space.
How are you integrating AI into your DevOps practices? I would love to hear your thoughts!

https://clickup.com/blog/how-to-use-ai-in-devops/

Nat-Gateway vs Nat-Instance

Basel Mohamed Alam — Thu, 25 Jul 2024 10:22:57 +0000

These services allow resources in private subnets to access the internet but have some key differences. Let's break this down:

Managed Service vs. Self-Managed:
- NAT Gateway: Fully managed by AWS.
- NAT Instance: Self-managed EC2 instance that you configure as a NAT.
Availability:
- NAT Gateway: Highly available within an Availability Zone (AZ). AWS automatically scales it as needed.
- NAT Instance: Single point of failure unless you implement your high-availability setup.
Bandwidth:
- NAT Gateway: Can scale up to 45 Gbps.
- NAT Instance: Depends on the EC2 instance type. Generally lower than NAT Gateway.
Maintenance:
- NAT Gateway: No maintenance required.
- NAT Instance: Requires patching, updating, and managing the EC2 instance.
Cost:
- NAT Gateway: Charged based on the number of NAT Gateways you use and the amount of data processed.
- NAT Instance: Charged based on EC2 instance pricing, which can be cheaper for low-volume workloads.
Security Groups:
- NAT Gateway: Cannot be associated with security groups.
- NAT Instance: Can be associated with security groups for additional network control.
Bastion Server:
- NAT Gateway: Cannot be used as a bastion server.
- NAT Instance: Can be used as a bastion server for SSH access to private instances.
Port Forwarding:
- NAT Gateway: Does not support port forwarding.
- NAT Instance: Supports port forwarding.
IP Addresses:
- NAT Gateway: Comes with a static public IP address.
- NAT Instance: You can use Elastic IP addresses, which you can bring your own, or use AWS-provided ones.
Performance:
- NAT Gateway: Optimized for NAT performance.
- NAT Instance: Performance depends on instance type and network configuration.

In general, NAT Gateway is recommended for most use cases due to its scalability, availability, and ease of management. However, NAT Instance might be preferred in scenarios where:

You need fine-grained control over the NAT device.
You're looking to minimize costs in low-traffic environments.
You need features like port forwarding or using the NAT as a bastion host.

Connect External Public IP to Private VPC on AWS

Basel Mohamed Alam — Thu, 11 Jul 2024 22:15:11 +0000

Having a Private AWS VPC and Want to Connect Public External Ip to it

takeing into consideration to Configure the Following:

VPC CIDR BLock
Routing Table
ACL
Security Group
and using IAM policy to Specifiy the Ip as Well.

Tried to Configure these Steps but with no luck, and couldn't ping from inside EC2

Migrating from Gitlap to Github enterprise

Basel Mohamed Alam — Thu, 27 Jun 2024 04:34:35 +0000

when migrating from Gitlap to GitHub enterprise, having more than one organization, projects and repos, and many users what is the best approach to achieve successful migration with the least operational or scripting effort and having the same structure for everything?

tried searching for an approach but found all answers that adding it one by one or by a third-party tool, expecting that would copy resources in bulk

Is there any AWS Service that can help achieve this process faster??

Serverless Frameworks: Optimizing Serverless Applications

Basel Mohamed Alam — Tue, 18 Jun 2024 04:19:59 +0000

Serverless computing has revolutionized the way we build and deploy applications, offering scalability, reduced operational overhead, and cost-effectiveness. Function-as-a-service (FaaS) and serverless frameworks are at the heart of this transformation, enabling developers to focus on writing code without worrying about managing servers. In this post, we'll explore how to optimize serverless applications using various open-source serverless frameworks and tools.

Understanding FaaS and Serverless Frameworks
Function-as-a-Service (FaaS): FaaS allows developers to execute code in response to events without the need to manage server infrastructure. Popular FaaS providers include AWS Lambda, Google Cloud Functions, and Azure Functions.

Serverless Frameworks: These frameworks simplify the deployment and management of serverless applications. They provide a structured way to define functions, events, and resources, and handle the underlying infrastructure for you. Key serverless frameworks include:

AWS Chalice: A Python framework for building serverless applications on AWS Lambda and API Gateway.
Claudia.js: Simplifies deploying Node.js projects to AWS Lambda and API Gateway.
OpenFaaS: An open-source framework that allows you to deploy serverless functions on any Kubernetes cluster.

Whether you are just starting with serverless or looking to fine-tune your existing applications, these strategies and tools will help you make the most of your serverless architecture. Happy coding!

Feel free to share your thoughts and experiences with serverless optimization in the comments below! Let's learn and grow together in the world of serverless computing.

Serverless #FaaS #AWS #CloudComputing #DevOps #Optimization #Monitoring #OpenSource #Programming #DevTo

OpenLambda: An open-source serverless computing platform for running functions written in any language.

VMware price hike forces Australian company with 24,000 virtual machines to jump ship

Basel Mohamed Alam — Sun, 26 May 2024 14:54:51 +0000

As I mentioned in this post-https://lnkd.in/dp-U7yPD - 2 months ago
that we are facing a new era of migration projects
this will be a tsunami from VMware to other competitors or open-source virtualization, In the recent Nutanix Next conference in Barcelona, Computershare CTO Kevin O'Connor, the company will likely abandon VMware's hypervisor soon to focus exclusively on Nutanix products.
sources:

https://lnkd.in/dHKpcNdt