Projects

A curated selection of projects that showcase my expertise in full-stack development, design systems, and devops.

Sketch Flow — A Sketching Tool

  • Architected and built a real-time collaborative whiteboard system supporting multi-user concurrent editing with low-latency updates using WebSockets.
  • Implemented a custom HTML Canvas rendering engine from scratch (no third-party drawing libraries), supporting shapes, freehand drawing, text, erasing, and interaction state management.
  • Designed a scalable real-time event pipeline using Redis as an in-memory queue, enabling fast UI updates while asynchronously persisting canvas state to PostgreSQL.
  • Developed invitation-based access control for shared canvases, allowing secure multi-user collaboration.
  • Optimized rendering and network payloads to handle high-frequency drawing events without frame drops or UI lag.
  • Built a monorepo architecture using Turborepo, sharing types and utilities across frontend, HTTP API, and WebSocket services.
  • Designed a fault-tolerant backend using Express and WebSockets, capable of handling multiple concurrent sessions per canvas.
  • Modeled relational data using PostgreSQL + Drizzle ORM, ensuring consistency between real-time state and persistent storage.
  • Containerized services using Docker to ensure reproducible development and production deployments.
  • Delivered a production-grade feature set comparable to Excalidraw, including free real-time collaboration.
Next.jsNode.jsExpress.jsRedisWebSocketsTurboRepoDockerShadCnReact.jsTypeScriptDrizzlePostgres SQLTailwind CSS

Insta-Transcribe — Vertical Video transcribing platform

  • Developed a Next.js application enabling users to seamlessly upload vertical videos (Instagram Reels or YouTube Shorts) to an AWS S3 bucket, integrating AWS Transcribe for automatic transcription.
  • Implemented a robust AWS S3 storage solution to efficiently store both the original videos and their corresponding Transcribe-generated transcription files, facilitating easy retrieval and management.
  • Utilized ffmpeg wasm to dynamically add captions derived from the transcription files to the original videos, enhancing accessibility and user engagement.
  • Enabled users to edit captions and timestamps directly within the application, providing a user-friendly interface for customization and improving overall user experience.
  • Incorporated advanced features allowing users to customize caption text color, offering a personalized touch to the videos and enhancing visual appeal.
Next.jsNode.jsAWS S3AWS TranscribeWASMZustandFFMPEGZustandShadcnTailwind CSSprismapostgress SQL

Learnify - Platform for Educators

  • Developed Learnify, an innovative web platform using Next.js and a custom frontend component library Shacdcn UI, empowering educators to create and share courses seamlessly.
  • Implemented secure user authentication using Clerk, enabling teachers to upload video chapters, attachments (PDFs), and manage their courses with Prisma ORM and Postgres SQL.
  • Integrated Stripe payment system, providing users with a smooth purchasing experience, enhancing monetization for course creators.
  • Utilized MUX player for an engaging video playback experience, making it effortless for learners to access course content.
  • Employed Zustand for efficient state management, ensuring a responsive and dynamic user experience.
  • Adopted modern technologies like Tailwind CSS and Zod for enhanced UI/UX and robust data validation, creating an appealing and user-friendly environment for educators and students.
Next.jsNode.jsShadcnTailwind CSSZustandclerkprismapostgress SQLzustand