KASI VINOTH S
DATA SCIENTIST

Hello, I'm a seasoned Data Scientist with a unique blend of Nanotechnology and Computer Science expertise. Currently, I'm on a quest to develop super-efficient algorithms that optimize business problems with the perfect predictive modeling.
    My skills are honed in Text Analytics using deep learning techniques, and I'm always on the lookout for innovative ideas and logic that add to my understanding of NLP and language itself. I'm a Knowledge Graph semantic Technology whiz, extracting insights from texts like nobody's business.
    Lately, I've also fancied myself as a Computational Linguist, although I don't have an official degree to back it up. Language is fascinating, and I believe patience is the key to unraveling its mysteries. When I'm not pushing the boundaries of AI, you might find me honing my chess skills or chatting with the ever-amazing ChatGPT - seriously, have you tried it?

PROJECTS EXHIBIT:

BANKRUPTCY PREVENTION
deployed on:
https://bankruptcy-prevention.herokuapp.com/

This web app predicts whether a company is getting Bankrupted or Not Bankrupted based on some key factors. ANN model is used with keras tuner for hyper parameter optimization

SENTIMENT ANALYZER
deployed on:
https://sen-analyzer-log.herokuapp.com/
DOCKER repo for the app :docker_link

This web app can predicts the sentiment of the user about a product. The training is done based on Amazon Realme Headphones reviews. The analyzer have also been experimented with simple LSTM model.
A docker image was made for the sentiment analyzer Flask app.

CAT-DOG CLASSIFICATION
Not deployed yet.
repo link: https://github.com/Vinoth-24/Cat-or-Dog-Classification.git

Using VGG16 CNN model we use Transfer learning technique to predict whether a given image is a Cat or a Dog.

FOOD SUGGESTION APP:
deployed on:
https://food-suggestion-app.herokuapp.com/

A Food suggestion algorithm was made for an Indian restaurant (250 recipes) with the help of K-means clustering technique and intensive eda analysis.

Deployment of your model on Heroku web using Streamlit library
article link : click here

This article is a procedural guide for beginners to start deploying their ML models into Heroku web using Streamlit library.
SKILLS

Programming Languages & Tools:
Python, SQL, SPARQL, GIT, LINUX, VMware, Google Colab, Pycharm, Jupyter Notebook, VScode, Bash, Tableau, MS Excel, DBweaver, Docker Desktop, CLI
Analytics & Algorithm Development:
I'm skilled at defining tasks, strategizing with my team, and managing projects. I'm adaptable, which allows me to deliver results seamlessly without any disruption.
Basic Data Science Skills:
I have experience in data visualization, feature engineering, data modeling, text mining, supervised and unsupervised model building, exploratory data analysis, data scraping, data cleaning, and transfer learning.
NLP-related Skills:
NLP-Related Skills:As an NLP practitioner, I specialize in text similarity analysis, knowledge extraction through sentence structure parsing, and category-based information extraction using techniques like Regex and Hearst patterns. I also have extensive experience building and fine-tuning models for identifying named entities, with a particular focus on domain-specific tasks. Some of the models I've worked with include T5, flan T5, ESG Bert, and environmental claims.
My expertise in NLP allows me to take on a variety of tasks, including Q/A, question generation, OCR, text summarization, SNLI, chatbots, speech recognition, and token and text classification. Additionally, I have a strong grasp of statistical similarity measures and use them to classify and understand documents based on their content. For more complex tasks, I use statistical embedding-based approaches for entity linking and co-referencing.

Education

Ever since I was in high school, I have been captivated by Mathematics and Computer Science. My fervor for programming and statistics led me to pursue Engineering in Nanotechnology, where I had the privilege to work alongside some remarkable scientists in India. Unfortunately, due to the pandemic, my plans to pursue a PhD were put on hold, and I returned to my roots in Computer Science, rekindling my passion for Data Science.

I enrolled in a comprehensive Data Science course at ExcelR institute in Bangalore, where I honed a diverse range of skills (mentioned above). Since then, I have been determined to enhance my skills and knowledge by taking various courses and continuously challenging myself. I aim to find motivation and inspiration through meaningful connections with others that push me towards achieving my goals.

Experience

   As an avid lover of nanotechnology, data science, and NLP, I have spent the past few years geeking out over the latest trends and developments in these exciting fields. My internships in nanotechnology gave me a chance to work with some pretty tiny materials, but the experience was huge in terms of expanding my knowledge.
    At Trell, I dove deep into the world of data and was thrilled to learn about the powerful capabilities of Advanced Excel and Python. I found myself excitedly crunching numbers and tinkering with various models in my spare time, which definitely improved my skills.
    Working with ML models at AiVariant was an adventure in itself - I felt like I was in a techy sci-fi movie as I experimented with different neural networks and tackled various classification problems. It was a real thrill to work on something so cutting-edge and forward-thinking.
    And now, at LeadSemantics, I'm part of a team that's focused on extracting knowledge from texts using some seriously impressive LLMs and semantic tech. I've learned so much about the value of language quality and how it can be applied to so many aspects of our lives. Our CEO has been a fantastic mentor who's encouraged me to explore different ideas and technical approaches, and it's been truly rewarding to work with and have mentored such a talented group of people.
    While I take my work seriously, I also believe that having fun is important. In my free time, you can often find me getting nerdy with the latest NLP research or chatting with other professionals to learn more about their perspectives. I believe that staying curious and open-minded is key to continued growth and development in any field.
    So if you're looking for someone who's enthusiastic, curious, and a little bit geeky, then look no further. I'm always eager to explore new opportunities and connect with other professionals who share my passion for data science, and NLP.

Feel free to reach out to me through mail - vino24995@gmail.com or phone-+919953253902.
Site under Progress....