Hi, I’m Varun Prajapati
a Data Scientist

I hold a Master's Degree in Cyber Security from NIT K. I also have a Master's in Data Science from Liverpool John Moores University. As a solutions expert, I am in charge of complex data sets belonging to financial and crypto markets. You can trust me to forecast trends and make sense of the data. In personal times, my friends tell me, I am fun to be with. I like science fiction novels and movies. My likes include indoor and outdoor activities altogether. I love discussing anything that catches my fancy - right from current affairs to rocket science!

Personal Portfolio Images

My Skill

  • skill
  • skill
  • skill
  • skill
  • skill
  • skill
  • skill
  • skill
  • skill
  • skill
  • skill
  • skill
    EC2
  • skill
    S3

My Resume

Academics

M.S. in Data Science

Liverpool John Moores University (2020 - 2022)
Merit
  • Implemented a Cryptocurrency price predicting model using Long Short Term Memory (LSTM)
  • Compared Single and Multiple Cryptocurrency Multivariate Price Prediction Models
  • Correlated different reasons of Price predictions based on historical price actions and events.
  • Tested Different approaches like LSTM, ARIMAX, Holt Winter’s Exponential Smoothening and NARDL.

P.G.D. in Data Science

International Institute of Information Technology – Bangalore (2020 - 2021)
3.75/4.00
  • Specialization in Natural Language Processing (NLP)
  • Developed and presented 10 Data Science Projects in different sectors
  • Implemented Logistic Regression for binary classification problem
  • Developed a Forecasting Model using Regression Techniques
  • Used open-source software rasa to develop a 2-level chat-bot

M.Tech in Cyber Security

National Institute of Technology – Kurukshetra (2017 - 2019)
8.53/10.00
  • Specialization in Smart Card Security Systems
  • Outcome of research was presented and published in 2 international conferences
  • The research was an outcome of the R&D works undertaken under the YFRF project under Visvesvaraya Scheme of Ministry of Electronics & Information Technology, Government of India.
  • Implemented different cryptographic algorithms from scratch like AES, DES, RSA, Diffie-Hellman and hashing algorithms like SHA, MD5.
  • Created, implemented and tested vulnerabilities of user authentication schemes with user anonymity for military and healthcare domains

B.E. in Computer Science

TCET – Mumbai University (2013 - 2017)
8.37/10.00
  • Research was done in cryptographic algorithms
  • The outcome of this research was published in Tata Mcgraw-Hill Journal of Computer Science 2017
  • Compared existing cryptographic algorithms used for key sharing and data encryption
  • Created and implemented an hybrid algorithm which supported multilayer encryption and on the fly password generation using cursor movements
Features

Technologies

Python
80%
MySQL
70%
Rasa
75%
Django Backend
80%
Apache Server
70%
Git
80%
Zyte
80%
Shell Scripting
70%
Linux Server Management
80%
AWS S3
60%
AWS Instance Management
80%
Features

Applications

Web Scrapping
70%
PDF Scrapping
60%
ETL
80%
Machine Learning
60%
Data Visualization
55%
Database Optimization
50%
Web Development [Backend]
70%
Automation
70%

Job Experience

Squarepoint Capital

Software Developer (2022 - Present)
  • Worked on different aspects of data gathering and processing from vendors.
  • Worked on different inhouse frameworks to optimize existing systems.

Bloombytes Private Limited

Python Developer (Financial Analyst) (2019 - 2022)
  • Built data scrapping solutions to get daily data from 100+ firms
  • Extracted and Automated data extraction from PDF’s and XML’s
  • Studied and derived analysis from different SEC forms and filings like NCEN, N-PORT, etc.
  • Extracted and analysed data from different public financial data sources like ADV and FINRA
  • Scraped data from various social-media platforms for professionals
Python Developer (ETL Expert) (2019 - 2022)
  • Automated Data Extraction and cleaning from different sources.
  • Extracted tabular data from PDFs using Tabula, Camelot and Pike PDF
  • Extracted data from XMLs and created an automated solution to load data in MYSQL DB.
  • Gathered Social Media data using Proxy networks
Python Developer (System Administration) (2021-2022)
  • Managed the security policies on AWS for Servers
  • Created Lifecycle policies, Snapshots, Image and Instance Management
  • Worked with AWS S3 and created system to do periodic file backups
  • Created Shell scripts and managed cron jobs for automation of tasks
Python Developer (Backend Developer) (2020-2021)
  • Designed and implemented Backend of a website
  • Core functionalities included a star based rating system, reviews, favourites and content and profile tailored for each user
  • Used tools like faker to populate database with spam placeholder contents during testing
  • Optimized the database performance by reducing total interactons and optimization techniques
Python Developer (Web Scrapping Specialist) (2021-2021)
  • Scrapped data from a dynamic website where data was loaded after javascript execution
  • Used Scrapy and Selenium for automation of web page browsing and data extraction
  • Scrapped data where captcha was present due to consecutive hits and proxy was not an option
Python Developer (Database Optimization) (2021-2021)
  • Studied the working of engines and indexes in detail for MySQL server
  • Optimized an already existing project by tuning system variables, modifying settings and trial and error approaches.
  • Studied the data in detail to understand the relation of tables and optimised slow queries by simplifying them

Projects

Data Science Projects

  • Forecasting model to Predict demand for shared bikes from historical trends
  • Fraudulent Transaction Detection from Bank Transaction details Dataset
  • Sales lead detection and retention strategy on Telecom company Dataset
  • RASA based chat bot for online delivery and dining company

Python Projects

  • Website blocker application which restricts access of customised website on windows for selected hours
  • Webmap which helps to group data on zoom and create a summarized view of geological statistical data
  • GUI based windows dictionary application from static dictionary sources

Cyber Security Projects

  • Multi-server Smart Card User Authentication Scheme
  • Single Server Remote User Authentication Scheme for Military
  • VoIP based Remote User Authentication Scheme
  • Implemented different cryptographic algorithms from scratch like AES, DES, RSA, Diffie-Hellman and hashing algorithms like SHA, MD5.
  • 2-factor Remote User Authentication for Telecare Medical Information System

Publications

Testimonial-image
Journal of Information Technology Research (JITR) - 2022 Edition

A Robust Authentication System With Application Anonymity in Multiple Identity Smart Cards

User Authentication plays a crucial role in smart card based systems. Multi-application smart cards are easy to use as a single smart card supports more than one application. These cards are broadly divided into single identity cards and Multi-identity cards. In this paper we have tried to provide a secure Multi-identity Multi-application Smart Card Authentication Scheme. Security is provided to user’s data by using dynamic tokens as verifiers and nested cryptography. A new token is generated after every successful authentication for next iteration. Anonymity is also provided to data servers which provides security against availability attacks. An alternate approach to store data on servers is explored which further enhances the security of the underlying system.

Testimonial-image
Journal of Neural Computing & Applications - 2021 Edition

Machine learning and smart card based two-factor authentication scheme for preserving anonymity in telecare medical information system (TMIS)

Telecare medical information system (TMIS) is used to connect patients and doctors who are at a different location from each other. The authentication of the user and system is very crucial as the medical data of the user is stored on the server. Many systems have been developed in order to achieve this goal. We show some vulnerabilities of existing systems in this paper. We then propose a secure authentication mechanism to achieve the same goal. Machine learning and the nonce-based system is used for authentication of the entity and to prove the freshness of transmitted messages. Smart card blocking mechanisms have been included in each phase of the proposed system to prevent unauthorized access of data. The proposed system has been evaluated formally with the AVISPA tool. Then the proposed model has also been checked against different attacks and evaluated for different functionalities. We provide relative analysis with some recently proposed models and show our proposed system is relatively more efficient and secure.

Testimonial-image
FTNCT - 2019 Edition

An efficient SIP Authentication Scheme for Multiserver Infrastructure

Session Initiation Protocol (SIP) is used between user and server to generate and agree upon a new session key through authentication mechanism. Many single server SIP authentication schemes have been proposed, but we need multiserver scheme to make it more realistic and feasible for implementation. It was found that some schemes work only for single server model and are not possible to scale up for multiserver environment. While some models support multiserver architecture but are vulnerable to known attacks. In this paper, we propose a new authentication scheme for SIP which can support multiserver environment. Security evaluation was done against various known attacks to check strength of the proposed scheme. Various functionalities like mutual authentication and user anonymity is also supported by the newly proposed scheme. Further our model is relatively more efficient as we establish connection only by using hash functions which are lightweight operations.

Testimonial-image
ICCES - 2019 Edition

Secure and efficient session initiation protocol authentication scheme for VoIP communications

Session Initiation Protocol (SIP) is widely used in Voice-over-IP (VoIP) communication to initiate communication between user and server. SIP authentication mechanism was found insecure against various attacks because of its inherent open text architecture. We analyzed different models for SIP authentication and noted their vulnerability against various attacks. Major identified weakness was lack of mutual authentication in various schemes. In this paper, we provide a new authentication model for SIP communication for secure communication establishment. We use AVISPA tool for formal security evaluation of our model. We then prove how our model can resist against various known attacks and provide mutual authentication. Our model is relatively more efficient as we use symmetric key cryptography with ECC instead of asymmetric key cryptography.

Testimonial-image
TMH - 2017 Edition

ASCII and Timestamp Based Encryption Algorithm

With the advent of internet of things, smartphones and high speed internet, the amount of data being transacted through small devices has increased exponentially. With this the risk of data being intercepted and modified in transit has also increased at the same pace. There is a need for encryption systems which can encrypt data to secure it while in transit. But this encryption system must be small enough to work on small devices which have much less computing power than a personal computer, and it must also be fast enough, so that the encryption does not turn into a bottleneck during data transfer. Apart from being fast and less resource intensive, the encryption system must also ensure that it can withstand the brute force and cipher text analysis attacks it will be subjected to on a regular basis. One such algorithm is proposed in this paper.

Contact With Me

contact-img

Varun Prajapati

Python Developer

If you have any questions feel free to contact me. If you don’t get answer immediately, I might just be travelling through the middle of nowhere. I’ll get back to as soon as possible.

Email: info@varunprajapati.com