Rupesh Keesaram

NLP | ML | Python


(per hour)


I am a NLP engineer with expertise in machine learning and python. As a freelancer I have worked on following projects:
  • Algorithmic Trading System
  • Tech Stack: Reinforcement Learning, Machine Learning, Python3
  • Description: This intelligent reinforcement learning system uses historical data and some ensemble techniques to make data-driven trading decisions, adapt to market fluctuations, and optimize trading strategies, all with the goal of achieving superior returns within the Nifty 50 market.
  • VocalVerse: SocialSpeak Manager 
  • Tech Stack: Deep Learning, Automatic Speech Recognition (ASR), Python3
  • Description: This is a cutting-edge speech-to-text model designed to revolutionize social media content management by generating accurate transcribes for a provided audio file. I’ve also created a pipeline which can seamlessly extracts audio from reels and translates in into captivating captions. This simplified the way we connect with our audience.
  • Layoff Insights Analyzer
  • Tech Stack: Selenium, BeautifulSoup, Python3
  • Description: This is an advanced crawler meticulously designed to navigate, extracting comprehensively layoff-related data. This tool effortlessly compiles the information into a CSV file, facilitating in-depth analysis. Complementing this, a dynamic Tableau dashboard is crafted, offering actionable insights and visualizations for a nuanced understanding of the evolving job market landscape.
  • Industry Projects
  • NLP (Natural Language Processing)
  • Automated the extraction of unstructured data and performed preprocessing using NLP techniques such as tokenization, stemming and lemmatization across various file formats.
  • Utilized Large Language Models (LLMs) to generate word embeddings and derive sentiment and risk scores for unstructured data originating from different reports.
  • ML (Machine Learning)
  • Conducting thorough hyperparameter tuning and model optimization to enhance model performance, while evaluating and selecting the best models using appropriate metrics.
  • Leveraging the capabilities of AutoML to automate the process of model selection, composition, and parameterization, thereby streaming the ML workflow for improved efficiency and effectiveness.
  • Interactive Data Analysis:
  • Developing engaging & interactive 3D plots using Plotly, enhancing data visualization capabilities for in-depth analysis.
  • Creating dynamic and user-friendly interactive dashboard application leveraging various features of Streamlit, allowing seamless navigation and exploration of data insights. Technical Skills:
  • Programming & Languages : Python3, Java and C
  • Machine Learning : Regression and Classification, NLP, LLMs
  • Data Analysis & Visualization : Pandas, Numpy, Matplotlib, Seaborn, Plotly, pandas-profiling etc.
  • Big Data Technologies : Spark, Flink
  • NLP : Spacy, NLTK, LangChain
  • Web Development : HTML 5, CSS 3, Bootstrap 5, Angular 8, Flask, Streamlit
  • Database : MySQL, Cassandra
  • Visualization Tools : Tableau
  • DevOps & Build Tools : Git, Dockers, Jenkins, Maven, Gradle
  • Cloud Platforms : AWS and Heroku


NLP Engineer

  • March 1, 2022 - continue

Project 1 : Credit Risk Modelling (AFC) Description : This is a ML-NLP fusion project where the bank provided different information like financial details, Risk Reports, Lenders Report, Production details of different their clients. We were asked to develop a robust model which can predict the credit default risk rate from 1 to 10, whereas 10 being the severe risk and 1 being the lowest risk. rn rnProject 2 : Providence Description : Providence is a platform for data scientists and users where they can perform AI/ML operations with ease. The operations may include training different problem specific models, model evaluation, visualizing the data in real time, making predictions etc. Responsibilities :Developing features like preprocessing pipeline, feature engineering pipeline, modelling pipeline and creating APIs Doing R&D on different charting libraries in react environment & data loading. Developing charts from scratch using D3.js

Report Freelancer

Ready To Get Started

QWIRK is a brand owned by Gigart Solutions Incorporation, Delaware, USA. The company is a freelance marketplace to help firms find quality professionals when needed, to be a part of your workforce.