Atharv Naik />
Web Developer
AI/ML
~$ ./about.sh
Hello! 👋 I'm an undergrad student at IISER Kolkata, majoring. I love building scalable backend systems with Django and Flask. Also, passionate about AI/ML and eager to apply knowledge in real-world projects!
Experience />
Django Developer Intern
- SMC Backend upgradation project under the guidance of Dr. Dwaipayan Roy, CDS Department IISER Kolkata
- Upgraded the Mess SMC API, incorporating endpoints for student usage analytics
- Payment gateway integration using Stripe API and handling account balance and transaction history
- Implemented endpoints to generate and email PDF Statements of Account.
- Contributed to the backend design and implementation for seamless credit and debit processes
- Writing Unit Tests for the API endpoints
April 2023 - Present
Data Science Intern
- Summer project on Audio Data Analysis of Speech Data under the guidance of Dr. Haroon Rashid Lone, SIRL lab, Department of EECS, IISER Bhopal.
- Collaborated with a multidisciplinary team working on video, audio, and text aspects of the project.
- Conducted data analysis on a video dataset of 119 participants to predict social anxiety using their speech data.
- Developed a feature extraction pipeline to prepare data for analysis.
- Established a collaborative environment by setting up a Samba server for concurrent access to participant data by multiple teams.
- Explored and selected relevant audio features for emotion detection, focusing on anxiety.
- Implemented noise reduction techniques, including low-pass filters and spectral grating, to enhance audio quality.
- Trained Random Forest and SVM classifiers on the extracted features, achieving an F1 score of 0.85.
May 2023 - July 2023
Projects />
PlayTube: Video Sharing Web App in DjangoView Deployment
- Built and deployed a video streaming web app in Django
- Developed video upload, transcoding, and streaming using Celery workers and Amazon S3 storage.
- Built custom video player for HTTP Live Streaming of videos.
- Integrated user authentication with social login (Google) via django-allauth.
- Enhanced user experience with custom HTML email notifications and watch history for video resuming.
- Achieved efficient content delivery through ABS streaming and Redis queue for asynchronous transcoding.
Detection of Blue Whale A-Calls from Audio Signals
- Trained a CNN model to classify blue whale calls into an A-call or not
- Extracted audio features from 26 thousand audio files for training
- Deployed the trained model in a Docker container with Flask backend
Chess Engine in Python
- A chess game implementation in Pygame library
- The algorithm detects Checks and checkmate
- Used Stack Datastructure to implement the undo/redo move feature
- Game can be saved and resumed later