Anubhav Sarkar

Worked as a Data Scientist for two and a half years solving some challenging problems in the field of NLP. Experience in working with different Machine Learning algorithms like Linear Regression, Logistic Regression, SVM, XGBoost and Deep Neural Networks (CNN, LSTM) using Python, Keras and Tensorflow. I have mainly worked on NLP projects at my current company. 
Possess intermediate level of experience with visualization tools like Tableau and Apache SuperSet 
and Plotly (using Python). Also have some experience in working with Docker containers and building 
rest APIs using Flask and FastAPI on an Ubuntu linux server. 

Data Scientist | Express Analytics | October 2020 - April 2023
· Sentiment Analysis on Customer Reviews : Developed a Deep Learning based
Sentiment classification model (Flair Library) using customer reviews data. The
model was trained on 235,310 reviews with “Positive”, “Neutral” and “Negative”
as the classification labels. An accuracy of around 92% was achieved on the test
set.
· Emotion Analysis on Customer Reviews : Developed a Deep Learning based
Emotion Classification model (Flair Library) using customer reviews data. The
model was trained on 235,310 reviews with “Joy”, “Trust”, “Surprise”, “Anger”,
“Fear” and “Sadness” as the Emotion classes. An accuracy of 86% was achieved
on the test set.
· Topic Extraction from Customer Reviews : Used Non-Negative Matrix
Factorization (NMF) technique to extract relevant Topics from user reviews data.
Coherence Scoring was used to decide on the optimal number of topics that can
be extracted from the given set of reviews.
· Creating Automated Insight Statements : Using the Topic and Sentiment Model
results, create automated Insight Statements, so that report generation becomes
faster.
· Exploratory Data Analysis on Customer Contacts Data : Received Customer
Contacts data from client Belair which had around 17 Million records. Performed
EDA on the different available features including NULL value analysis and faulty
records detection. There were 11 attributes present in the dataset including age,
contact-number, email, traveller details etc.
Data Science Intern | Sensight Technologies Private LTD. | January 2020 –
October 2020
· Tracking Car Route Deviation : Created an algorithm for Detecting Deviation in
Car Routes to ensure lesser chances of accidents and assign Route Familiarity
Scores to the drivers based on their past trip details. Higher Familiarity Score
means lesser chance of getting lost and ensures safer driving habits.
· Detecting chances of accidents based on Car Video Analytics : Used SSD
MobileNet object detection algorithm along with CSRT Tracking algorithm to
detect and track objects close to the vehicle in question and generate alert
accordingly.

M.Sc in Big Data Analytics | June 2018 – May 2020 | Ramakrishna Mission
Vivekananda Educational and Research Institute | Howrah, West Bengal
o CGPA (based on 4 semesters) : 9.22 / 10