Full Stack Development
⚡ Building responsive website front end using React
⚡ Creating application backend in Node, Express, FastAPI & Django
DataScience & AI
⚡ Developing highly scalable production ready models for various deeplearning and statistical use cases
⚡ Experience of working with NLP projects
⚡ Complex quantitative modelling for dynamic forecasting and time series analysis
Cloud Infra-Architecture
⚡ Experience working on multiple cloud platforms
⚡ Hosting and maintaining websites on virtual machine instances along with integration of databases
⚡ Setting up streaming jobs from DB to Server or vice-versa on GCP and AWS
Education
Indian Institute of Information Technology BOMBAY
Master's in Physics 2021 - 2023
S.S.Jain subodh college ,University Of Rajasthan
Bsc in Physics Honors 2018 - 2021
Portfolio Repository
Experience
Completed 8 Jobs on upwork with 5 start rating
Build websites for clients using reactjs, nodejs and streamlit
Developed backend using Nodejs, FastAPI and host them on azure and aws.
Convert python scripts into streamlit application.
Developed a pipeline for retrieving the documents from ElasticSearch (contextual DB) and Pinecone (vector DB).
Did benchmark between Elasticsearch, Pinecone, and Milvus Databases on different types of pdfs and docs.
Implemented AWS cron lambda for fetching data from SQL, dividing it into chunks, and sending it into the SQS queue.
After retrieving the documents use the Roberta Model to get the answers to the questions based on the query.
Developed an Auto-completion function for the given data to receive the contextual-related tags from ElasticSearch.
Developed APIs for serving this functionality to the client to create, delete, update, and search the tags using FAST API.
Developed Twitter Sentiment lambda for fetching tweets, then filtering and sending to DynamoDb.
Technologies Used: Amazon Web Services (AWS) · AWS Lambda · Amazon S3 · Amazon Dynamodb · backend · Natural Language Processing (NLP). Fast API · Elasticsearch · Milvus · Pinecone
Maintaining the front-end and back-end design of the client’s website. Create react tables in the front end and change the data flow in these tables using SQL queries.
Developed Web pages using the SiteJet development tool.
Fix bugs and provide new solutions to the errors. Technology Used: React js, Nodejs, Site J
Projects
Apartment Website
- Scrap appartment data from other wetsites selenium and beautiful-soup
- Design Backend of this appartment sale website using Nodejs
- Developed the frontend using ReactJs
- I delivered this webapp for a client on upwork.
Resume parser and Filter by Job description
- Developed resume parser using Name Entity Recognization model and spacy to extract name, email and phone etc.
- Train the NER model after annotating some resumes and existing data.
- Filter the resumes based on a particular job description using semantic and contextual similarities.
- Developed a Django App to use these functionality.
Blog WebApp
- Designed Fronted in React then connected to the backend.
- Develop Backend in Nodejs and MongoDB to write a blog.
Book Recommendation System
- Developed a recommendation system using KNN (K nearest neighbor) algorithm on user ratings and book data.
- Designed the Frontend by Streamlit (python Library). This web app is hosted on Heroku Cloud Server.
E-commerce web application (MERN)
- Built a live E-Commerce web app hosted on the Heroku cloud. And also create an Admin panel for admin operations
- Developed APIs for Sign In/Up, user profile, admin panel cart, and CRUD functionality for products,users, and admin.
- Technology used: ReactJs, Nodejs, expressjs, and MongoDB.
Matrix Converter| (Web Development)
- Developed a website for converting second or third-order matrix into Jordanized forms using Python, web template.
- Many forms of the matrix are visible on the web page like Transpose, Square, Jordan Form, and Determinant.
My portfolio
- I developed a nicely framed website to showcase my portfolio.
- The webpage has the educational background page, projects, blogs, work experience, and contact details. Technology used: ReactJs, Html, CSS.
One Class Text Classification
- Our goal is to classify the given text into two classes but we have only one class in training data.
- We used ensembles of models and implemented a research paper to solve this problem.
Detecting contradiction and entailment text
- This is an NLP problem which defines that How two sentences are related. These sentences are entailed, natural and contradicted. Implemented an ANN after the embeddings from the BERT Model. Achieved 92.85% accuracy score
- Developed a TensorFlow-based model using BERT Model and ANN. The model has trained on the sentences dataset.
Open Source Contributions
Contributed Organizations
Contributions
Skills
Web Technologies
Databases Technologies
Cloud Services
AI/NLP
