Himanshoo Thakre
Full-Stack Developer
Building production systems with thoughtful architecture. I solve complex problems through clean code and pragmatic engineering decisions.
Selected Work
Production systems built with focus on reliability, performance, and maintainability.
Problem
CDC placement/internship notifications were posted on ERP noticeboard with significant delays, causing students to miss important opportunities. Manual checking was time-consuming and inefficient.
System Architecture
Automated notification system with Python scraper running as cron job, Gmail API for ERP session re-authentication, Next.js frontend for real-time notice display. Scraper posts notices.json to Next.js app for instant rendering.
Data Flow
Key Trade-offs
- →Used cron jobs for scheduled scraping instead of real-time webhooks (ERP limitation)
- →Implemented Gmail API for automatic re-authentication when ERP session expires
- →Chose Next.js for fast static generation and real-time data updates
- →Stored notices as JSON for lightweight data transfer between scraper and frontend
- →Balanced scraping frequency to avoid rate limiting while maintaining freshness
Outcome / Impact
Deployed to production with 300+ active users. Eliminated notification delays, ensuring students never miss placement opportunities. Became the go-to platform for campus recruitment updates.
Problem
Need for a lightweight, user-level thread library supporting creation, termination, and synchronization without relying on existing pthread implementations.
System Architecture
User-level thread library using clone() system call with support for thread creation, join, termination, mutexes, and barriers. Implemented thread-safe synchronization using TIDs.
Data Flow
Key Trade-offs
- →Used clone() system call directly for fine-grained control over thread creation
- →Implemented custom synchronization primitives instead of relying on kernel-level locks
- →Designed thread termination routines (foothread_exit) for clean joinable thread management
- →Validated library with tree recursion tests for correctness in parallel and sequential execution
Outcome / Impact
Successfully implemented a fully functional thread library with mutex support, preventing deadlocks and ensuring thread-safe concurrent execution.
Problem
Need for an intelligent search engine to efficiently retrieve relevant questions from large datasets with high precision.
System Architecture
Search engine using Term Frequency-Inverse Document Frequency (TF-IDF) algorithm for ranking, automated web scraping for data collection, dynamic Flask-based interface with HTML/CSS/JavaScript frontend.
Data Flow
Key Trade-offs
- →Chose TF-IDF over simple keyword matching for better relevance ranking
- →Used Selenium for dynamic content scraping where BeautifulSoup was insufficient
- →Implemented Flask backend for lightweight, fast API responses
- →Optimized search algorithm for handling 2400+ problems from LeetCode efficiently
Outcome / Impact
Achieved reliable, high-precision data extraction at scale with an optimized search interface enhancing user experience.
Work History
Professional experience building and scaling production systems.
Key Responsibilities
- →Increased subscription conversion by 35% by integrating PayPal Gateway via REST APIs, ensuring seamless transactions and security
- →Boosted visibility and lead generation by 15% by developing a marketing website using ReactJS, hosted on Hostinger with optimized SEO
- →Improved efficiency and accelerated task release by 20% by implementing I/O-driven plugins on monday.com, integrating Google apps in monorepo architecture
- →Enhanced workflow efficiency by deploying embeddings-based plugins on monday.com, integrating Google apps in monorepo architecture
Impact & Achievements
- ✓35% increase in subscription conversion through PayPal integration
- ✓15% boost in visibility and lead generation with SEO-optimized marketing site
- ✓20% faster task release with improved workflow automation
Key Responsibilities
- →Increased net load speed by 25% and strengthened UI consistency by refactoring and redesigning the dashboard using Angular and Ionic
- →Reduced server response time by 15% and enhanced core system performance by developing and optimizing reusable REST API endpoints
- →Lowered user drop-off rate by 20% and streamlined seamless onboarding by implementing OAuth for third-party login, boosting security
- →Ensured a reliable, scalable, and secure environment by deploying the application on Heroku, streamlining smooth deployment processes
Impact & Achievements
- ✓25% improvement in load speed with UI refactoring using Angular and Ionic
- ✓15% reduction in server response time through optimized REST APIs
- ✓20% decrease in user drop-off with OAuth implementation
- ✓Reliable deployment pipeline established on Heroku