RSD5095

EDUCATION

INDIAN INSTITUTE OF TECHNOLOGY BOMBAY (IIT BOMBAY)

Bachelor of Technology in Civil Engineering; GPA: 7.22/10 Jul ’17 – Aug ’21
• Awards and Honors: Special Mention Award ’19 (Hostel Organizational Award List, Hostel 3 Council)
• Certificates: Deep Learning Specialization – 5 courses (Stanford); Python for Everyone Specialization (Michigan)
TECHNICAL SKILLS
Languages: Java, JavaScript, Python; Frameworks: NestJS, Spring Boot, NodeJS, ExpressJS;
Databases: MySQL, PostgreSQL, Redis, MongoDB; Queue & Messaging tools: Kafka, Bull; Networks: Sockets
PROFESSIONAL EXPERIENCE
TECH4BILLION MEDIA PVT. LTD. (CHINGARI) | Software Development Engineer (Back-end) Oct ’22 – Present
Developed 2 full-scale services – Room-Feed & Private Call and optimized code for Live Room product using NestJS & NodeJS
 Room-Feed service – Developed feed service for Chingari app to integrate 3 products (video, audio and private call)
– Developed the code entirely in Redis and Kafka, utilizing Redis stack features such as RedisSearch and RedisJSON
– Successfully handled 2 Mn daily requests on top room APIs and 8 Mn daily requests on total service (latency < 20 ms)
– Aggregated geoData with Redis for showing videos nearby to user and developed a toggle to view private content
– Integrated 3 service API calls into 1 service API call to reduce the latency by 80% and optimize server load
– Implemented real-time follow-unfollow feature, enabling users to receive live updates on feed based on preferences

 Private Call – Designed architecture and databases for 1-1 private calling with socket connections for data transfer
– Utilized Bull Redis for scheduling tasks such as deduction of payment for private calls (every 1 min.)

– Implemented socket connections for services’ key features – call sending, receiving, declining and auto-repayment
 Live Room (1 Mn users) – Developed features of Live Room such as Battle, Poll/King of the room, Pause/Resume room
– Developed Battle features – Inviting users, user suggestions and winner declaration based on activities and gifts
– Integrated Telegram bot to effectively detect and handle NSFW content on the app; based on scores by ML models
– Developed a robust real-time leader-board feature within Battle and Poll functionalities using Redis sorted sort
– Implemented Pause room feature to ascertain duration based on live time, allowing users to join a room without login
 Chingari App Service – Executed updates to optimize services such as Music Search, API Rate Limiter and In-App banners
– Utilized Elastic Search to enhance Music Search feature; implemented Cron Jobs for real-time updates of new music
– Implemented App Banner feature using Node Cache and Redis, intelligently limiting Redis API call to 1/100 requests
– Employed IIFE (Immediately Invoked Function Expression) to efficiently load data into NodeCache upon service startup