Aasim Khan

Software Engineering Intern

Smart Search: Designed a smart search feature which understands the users queries, provided better property results and increased the app engagement by 2x. Android Front End: Designed parts of Android Application which help user filter the property results as per their requirement. Web Application: Conceptualized and created a website to be used as an alternative to the mobile application.

Android App Developer

Fastdeals Android App: Developed the flagship Android Application which connected to the APIs to carry out the business operation. Libraries used are: Volley, Picasso, GSON, Activeandroid.

Software Engineering

Restful APIs: Wrote node.js APIs which are required to carry out the business operations by the front end portal and the admin panel.

University of California, Santa Cruz

Credibility Score and Summarization of News Topics (FakeNet)

Designed a system to rate news articles on their credibility. Created a CNN based model which classifies news as Fake or Real. This model was trained on glove vector embedding of the news articles. The model achieved an accuracy of 97.8% on Kaggle Fake News Challenge. Summary of news topics was generated using Tex-Rank algorithm.

Online IDE

A simple and intuitive system that allows the users to code in their browser and execute their projects. Built on a node.js as the back-end. The system supports multiple features like directory management, code editor and code execution.

Smart Monitors:Eye Care System for PC Users Based on Blink Rate Prediction

Implemented an eye care system for computer monitors that involves tracking the blink rate of the user and alerting them if it goes below the norms. Eye regions are extracted using Haar Cascades classifiers. CNN model implemented using Keras is used to detect eyeblinks.

Stock Market Prediction using Technical Indicators

Prediction of stocks based on historical stock data. The systems train multiple machine learning models based on different stock market technical indicators and predict the values of the next N-days using the best one.

Virtual Stock Market Application

Created an Android Application with a node.js backend which simulated the actual workings of the stock market. Used as an educational tool to teach students about investing in the stock market. Linear Regression was used to predict the future stock prices based on current trends.

Medical Diagnosis Chatbot:

Developed a chatbot which takes in symptoms from the users and gives them a possible diagnosis. The chatbot is connected to a MYSQL database which contains all the symptoms and diseases.

Prediction of Online News Popularity

Designed a system which predicted the popularity of tweets posted by news twitter handles. Different features were extracted and a gradien b􏰀oosting based classifier was trained which provided an accuracy of 80%.