Sanjay Nithin

B.Tech in Information Technology

September 2020 - Present

• Machine Learning
• Cloud Computing
• Network Information Security

Higher Secondary School

June 2008 - March 2020

• 10th Grade : 95.8%
• 12th Grade : 89.5%
• C++ and Python Programming

eSomein Consultancy Pvt Ltd · Internship

Backend Developer

March 2023 - Present

• Handled the entire Backend (RESTful API) for a school-portal.
• Technologies used: Node.JS, Express.JS, JWT, Multer, Bcrypt, Mongoose.
• Tools used: Nodemon, Postman, MongoDB Atlas, Render Hosting Service, Firebase Storage.
• Tasks handled: Postman API documentation, Database Design, Render Hosting, Implementing Security Practices, using Firebase Storage for storing files, file handling, Payment Portal Integration using Razorpay SDK and Deployment in cPanel.

Visual Bloggers' Club VIT

Editorial Core Committee Member

May 2021 - Dec 2021

• Co-ordinated two events.
• Article about Self Development Industry and the reason for its popularity: Religion Of Self-Development.
• Article about how God might react to all that is happening around us: A Letter By God.

VITrendz

Tech member

May 2021 - Nov 2021

• Worked on my first collaborative project: GPA Calculator using vanilla.js.
• Technical writer for tech related posts on Instagram.
• Expanded my collaborative and creative boundaries.

Performance Analysis of a Micromodel-based Multinomial Classifier

2022 International Conference on Computational Science and Computational Intelligence (CSCI)

August 2022 - December 2022

• The paper focuses on analyzing the performance of overall model by creating seperate models for individual classes and aggregating them using a function.
• Four hundred resumes were collected from our university students and were parsed using Affinda Resume Parser.
• One-Hot Encoding was applied on extracted skills.
• Traditional machine learning supervised algorithms such as support vector machines, random forest classifier and unsupervised algorithms such as KMeans and Kmodes were developed.
• A single Feed-Forward Artifical Nerual Network was developed.
• A single Feed-Forward ANN was developed for each class and aggregated using a mathematical function.
• The performances of the developed models were studied and presented in the paper.

Securing Cloud File Storage using Blockchain

Ethereum, Solidity, Kaleido, Google Firebase, React, Node, Vercel, Render, Postman

Jan 2024 - May 2024

• The proposed architecture aims to enhance data security and access control in cloud storage environments through the integration of blockchain technology.
• By leveraging a blockchain ledger, the system maintains immutable records of users, roles, and encrypted files.
• Data Owners act as administrators, generating keys for various roles and encrypting files accordingly.
• Users access the system to obtain their keys, with Data Owners approving their access.
• Encrypted files are uploaded to the cloud, with associated Merkle roots stored in the blockchain alongside role and department information.
• When accessing files, the system retrieves blockchain records to verify user access permissions before decryption.
• Integrity checks involve comparing stored and recalculated Merkle roots.
• Only Data Owners possess the authority to delete files.
• This architecture ensures granular access control, auditability, and data integrity verification, mitigating risks associated with traditional cloud storage approaches.
• By decentralizing access management and leveraging blockchain's inherent security features, the system offers a robust solution for safeguarding sensitive data in cloud environments.

Phishing Detection Webapp

Machine Learning, Networking, Flask

June 2023

• Support Vector Machine, Logistic Regression, Random Forest Classifier, AdaBoost, and Ensemble models were trained on the dataset.
• The URL is taken as input, and all the attributes are manually calculated (refer to the source code).
• The attributes are then passed to the models for prediction.
• Flask, Jinja template engine, and minimal Bootstrap are used to display the results.
• Machine Learning packages such as scikit-learn, seaborn, pickle, numpy, and pandas are utilized.
• Python Networking and Web Scraping packages, such as BeautifulSoup, requests, and socket, are employed to calculate the attribute values.
• I completed the entire project, including building the ML models, evaluating the attribute values, and styling, on my own.

Covid Resources Sharing Webapp (Updated)

Node.JS, User Authentication, Template Engine

November 2022

• The webpages were served using Node.js and Express.js.
• The bcrypt module was employed for user authentication.
• Handlebars template engine, nodemon, Mongoose ODM, and cookies were utilized.

Cloud File Searching Webapp using AWS Kendra

AWS Kendra, AWS S3, AWS DynamoDB, AWS IAM

August 2022 - November 2022

• AWS services, such as Kendra, were utilized for searching through the contents of the file.
• The files were uploaded to an S3 bucket, and their hash values were stored in a DynamoDB collection.
• The S3 bucket was indexed by the Kendra service, which returned files containing the keywords.
• If a user uploaded the same document, the ETag (hash value) and file size were noted. If the values matched any record in DynamoDB, a plagiarism response was sent back to the user.
• The file server (fs) was used for file handling.
• User authentication methods, such as login and signup, were implemented. Passwords were hashed and compared using bcrypt.
• Express-fileupload was used for file handling.
• The implementation was scrapped due to costs issues, as Kendra is a very expensive service (costing around 80k per month for regular indexing).

Todo Desktop App

JavaFX, JDBC, Scenebuilder

December 2021

• MySQL and JDBC Connectivity were utilized in the development of this application.
• Scene Builder was used to design the interface.
• Users have the ability to add tasks, view tasks, and delete tasks.
• The app has a few flaws, such as relying on a local database. Tables need to be created in the console before the application can utilize them.

Covid Resources Sharing Webapp

Flask, Heroku Deployment, Jinja Engine, SQL, REST API

May 2021

• A Third-Party API was used to retrieve COVID-19 details.
• This application was created during a time when people shared information about resources on WhatsApp.
• Flask, Jinja engine, and SQLAlchemy were used in its development.
• Minimal Bootstrap was utilized for styling purposes.
• The application was deployed on Heroku when it still had a free tier.
• It was developed for a hackathon and was among the top five shortlisted teams.

MovieBuff Discord Bot

Discord.py, File Handling

Februrary 2021

• A bot was developed to provide information about films, film personalities, and maintain a watchlist within a Discord server.
• Initially, repl.it was used, and it was pinged every 5 minutes using a third-party service.
• The bot utilized a file (as a substitute for a database) to maintain the watchlist, which was shared across servers.
• Due to updates in the Discord.py module, the bot's intents were handled, and it was deployed on the Railway hosting service (service terminated).

Wikipedia Discord Bot

Discord.py, Wikipedia API

Februrary 2021

• A bot was developed to provide information about anything within a Discord server. This was my first project, and it helped me get well versed with Python environment.
• Initially, repl.it was used, and the bot was pinged every 5 minutes using a third-party service.
• Due to updates in the Discord.py module, I had to handle the bot's intents, and eventually, I deployed it on the Railway hosting service (service terminated).

Languages and Databases:

Frameworks:

Tools: