Skills
Languages: C++, Python, Javascript, SQL.
Libraries: TensorFlow, Djnago, AWS, NumPy, Pandas, Scikit-learn, matplotlib, seaborn, nltk, BeautifulSoup, Git.
Interests: Natural Language Processing, Recommender Systems, Deep Learning, Web Development, Algorithms and Data Structures.
Work Experience
-
Melopond Online Radio
Software Development Intern
- Engineered backend and songs feature extraction methods for the AI-based online radio station startup.
- Used AWS S3, Flask, MySQL, Collaborative filtering.
-
Samsung Research Institute Bangalore
Summer Intern
- Device Services Team: Developed a robust log files error classification and analysis system and received pre-placement offer from them.
- Used Python, MySQL, Pandas, Multiprocessing, Regex.
Publications
-
Personalized Diversification in Recommendation Systems: A Cluster Based Approach
Naina Yadav, Kartikey Singh, Anil Kumar Singh
- Worked with Associate Professor Anil Kumar Singh to develop a novel method to increase diversity in recommendations.
- Currently submission under review in Applied Soft Computing journal.
Projects
Here I present selected projects that I find quite interesting.
- Third Position in HCL Lucknow AI Hackathon
- Built a complete product under 24 hours working in a team of four to build a biased free source of news and used automatic topic assignment and ranking for it.
- Used Twitter APIs, Topic modelling(LDA), Flask, ReactJS.
- CODE
- Replicating-SeER
- Replication of SeER: An Explainable Deep Learning MIDI-based Hybrid Song Recommender System in TensorFlow2.
- Provided first publicly available copy of their dataset and code.
- CODE
- Vote for Change
- Created a polling website where users can poll on their topic of interest after signing up and shows statistics and graphs of the resulting poll to users.
- Part of the code.fun.do hackathon and used Django, jQuery for it.
- LINK
- Orthographic Languages Similarity Measurements
- Extracted similar words between Orthographic languages along with their distance by using provided corpora with the help of Longest Common Substring (LCS), n-gram and DICE algorithms.
- Part of CSE-443 (Natural Language Processing) coursework, received grade A for it.
- CODE
- Automated Direction Detecting Bot
- Created a wireless bot that processes an image and traverses according to the direction detected by it.
- Used VGG-19 architecture to develop the CNN model and part of final year project for bachelor’s degree.
- CODE
- Recommendation System
- Music and Movie recommender systems using collaborative filtering on public datasets.
- Used various matrix factorization algorithms and metrics to generate findings.
- Face and Emotion Detector
- Developed a face and emotion detection system capable of classifying various types of emotions by the usage of HAAR cascade classifiers and Fisherface algorithm built in OpenCV.
- This application also implements various emojis ,quotes ,colours and sound techniques for enhancing the experience.
- CODE
- WebTree
- An application which creates a PDF by using the structured graph of a webpage and it’s related webpages through the use of URL’s.
- It uses the connection between webpages to join them together and create an offline PDF of the webpage with the help of depth first search (DFS).
- CODE
Positions of Responsibility
These are the projects that I did as part of IIT (BHU) technical teams.
- Member of Technex-18 Technical Team
- Developed a scalable/modular web application for the tech fest of IIT(BHU) (used by more than 15k participants), solved various tech-related issues throughout the year
- Web Developer for Institute Day Portal
- Created a web application, working in a team for the Institute Day of IIT(BHU) used by 150 students at the institute to present their projects.