Rohit Negi rnegi4560@gmail.
com
www.linkedin.com/in/rohit-negi-b841b9246/
+91-8652108272
https://github.com/RohitNegi12
EDUCATION
A. P. Shah Institute of Technology, Thane (University of Mumbai) 2021– 2025
Bachelor of Engineering (B.E), Information Technology Honors with AIML
GEI's S. V. Joshi High School & Junior College, Dombivli (E) 2019 – 2021
Senior Secondary (XII), Science
INTERNSHIPS
● Completed AICTE Virtual Internship in Data Science Master through Altair
● Completed AICTE Virtual Internship in AIML Internship through AWS
● Completed TATA Data Visualization: Empowering Business with Effective Insights Virtual Experience
Program on Forage
TECHNICAL SKILLS AND INTERESTS
● Tools: Github, VS Code, Google Colab, JupyterLab
● Programming Languages: Python, C++, HTML, CSS, ML Libraries, DeepLearning
● Database: MySQL, PostgreSQL
CERTIFICATIONS
● Cisco PCAP: Programming Essentials in Python
● NVIDIA Deep Learning DLI
● Amazon AWS Academy Cloud Foundations
● Amazon AWS Academy Machine Learning Foundations
● Cisco CCNAv7: Introduction to Networks
PROJECTS
Multiple Disease prediction System
● Processed all the data taken from kaggle of brain tumor images, heart disease and diabetes
● For images I used sequential model and stacked convolutional layer, activation layer , pooling layer and
flattening layer from the tensorflow library, finally I applied dense network layer at last, I was able to
achieve 85% accuracy in prediction
● For the Other two models I used Support vector classifier and was able to achieve 77% test accuracy and
also I calculated and plotted the feature importance for both the diseases which will help know the impact of a
that feature on the disease which will help to know which features impacts more the disease
● UI was created with streamlit
Review Analysis Dashboard for Products
● On ecommerce platforms there are n number of products and each of them have many reviews after
scraping the reviews its very hard to manually go through them and analyze them, our system solves that
problem by automatically generating the graphs for the sentiments of each product and generates a word
cloud which represents the quality aspect that has been most talked about in the reviews
● The Dataset used for training the logistic regression model was Amazon fine food reviews
● The Data before training went through preprocessing steps the first step was tokenization, lemmatization,
stop words removal all this with the help of Nltk library and then word2vec technique used to represent the
words into vectors (numbers) using the Gensim library and finally trained the model and evaluated it, the
results were 80% accuracy
● Also used an open source model from hugging face platform which had better results from our model
because it uses transformers and it has been trained on large parameters.
● For Backend Flask was used and Streamlit was used for the UI.
ACHIEVEMENTS
Data Science Student Club Member | A. P. Shah Institute of Technology
Engaged in learning opportunities and participated in various knowledge training sessions. Assisted with accessing
program and institutional resources, and supported efforts in public relations and outreach
Participated in 3 Hackathons: participated in Smart India Hackathon, and my project was CCTV based
automatic crime detection, I learned about various technologies like 3D CNN’s (Convolution Neural Networks) ,
(2+1)D CNN’s learned about new researches and current of state of the art models like Tensorflow Movienets,
Unfortunately or prototype got rejected in the first round but I worked on it as it was a good topic and built a basic
version of it which can classify Anomalous Activities
My Next hackathon was of my university, the problem statement was to build a CRM system for HR’s by using
chatbots, we built the first version in first 12 hours and we were qualified for the final round, in that first version
the system was integrate with chatbot as when HR types any operation it was performed automatically on the
database I learned a lot of skills like integrating LLMs (Large Language models), Open source LLMs like
llama-3, mixtral etc
The first hackathon that I participated was by Data Science Wizard Organization and the problem statements was
about reducing Air Quality Index (AQI), I did the EDA analysis and built a model which could predict the AQI
which could help to detect whether the AQI is good or not for that particular region
