Computer Science graduate, seeking internship/co-op opportunities in the field of Software Development and Data Science. With an extensive well-rounded experience at 2 fast-paced startups i.e. Tracxn & Innovaccer.
At Tracxn, I developed the data mining model for the organization which enabled them to crawl new startups by their online presence. This model was based on spider crawling thus making it easier to aggregate and create a startup portfolio.
Currently I'm working as a Graduate Student Assistant at the University at Buffalo Information Security Office. My work includes network flow visualization and analytics along with anomaly detection in network using Machine Learning Models.
Apart from developing cutting-edge, robust software solutions for companies to scale effectively, I also have successful projects in the field of Machine Learning.
Looked after the network ingestion pipeline
Created storage and visualization of network flow using ElasticSearch, Kibana and Logstash
Working on anomaly detection in networks using Machine Learning
Looked after the backend of the platform and created massive online data mining models to crawl startups based on their online presence.
Migrated several webpages from grails to React.JS
Created a name domain crawler which gathered all the newly registered,deleted domains thereby enabling ~1 Million domains addition to the database everyday.
Optimised database queries thereby reducing the required servers to handle daily load
Created decision based models for various clients using Natural Language Processing and Statistical Machine Learning.
Major Contributor of Datashop Lens that is used for retail based clients.
Developed a SixDesk library in Python to manage SixTrack Simulations using local client database (SQLite) and a centralized database (MySQL) replacing the existing mechanism which uses shared filesystem (OpenAFS)
FootBall Analytics Dashboard Using Player TweetsDeveloped a web dashboard to visualize various analytics using d3 and angular. The data used was fetched from twitter for various players using Python and stored and processed in MongoDB for aggregation. The site is live and can be checked at http://ec2-34-210-46-217.uswest- 2.compute.amazonaws.com/football/
Question Answering System Based on Twitter DataDeveloped a Question Answering System based on Twitter Data as a graduate course requirement for Information Retrieval. Used natural language processing to extract relations which are used to answer questions. The site is live and can be checked at http://www.aneeshbhatnagar.com/n-quire
Simple DynamoDB on Android With Replication and Fault ToleranceDeveloped an Android Messenger which implements simple Dynamo DB along with fault tolerance and replication. Uses content provider and basic multicast for messaging. Recovering Device can quickly sync with its negihbours to get the missing data
Optical Character Recognition for Hindi Language using Neural NetworkDeveloped an OCR for Hindi language using backpropagation neural network in Python.
Classifying Ephemeral vs. Long Lasting Content on the WebApplied techniques of Machine Learning to build a classifier that is best capable of predicting how users would label the websites (evergreen or ephemeral).
MoviePie, Web Application
Seizure Detection in Intracranial EEG RecordingsApplied techniques of Machine Learning to develop a basic seizure detection system for Intracranial EEG recordings using Fast Fourier Transform and Linear Regression.
Twitter CrawlerConstructed a Twitter Crawler to crawl tweets for specified keywords/hashtags using both Search and Stream API. The tweets would be stored in MongoDB and processed to be indexed in Solr.
Used python, solr and mongodb