Abhishek - Abhishek Bhattacharya

Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

ABHISHEK

Seeking opportunities where I can utilize my analytical, mathematical and technical


skills to solve real life problems related to analyzing a big volume of datasets to draw
insights that can help with business decisions.

Contact:8825149142

Email:abhishek11.dx@gmail.com

Experience
Worked as data analyst for 2 years in incident management team in Vodafone. Analysis of KPI metrics
and TAT using MySQL and Excel to understand and excel on the performance based on the insights
USING data visualizations tools and take data driven decisions.

 Managed database service support for systems and applications by managing a team of 30+
members in multiple geographic locations especially UK to ensure that SLAs were consistently
met.
 Ensure the Incident Management process is conducted correctly and all the KPIs are met.
 Database based service query using SQL and later depicting it on various analytical
Tools such as Tableau,power bi helped us to prioritize and Diagnose the root cause
Analysis(RCA).,it helped to reduce 70% escalations related to user access management.

Skills

Programming Python, SQL

Visualization Tableau, Power BI, Jupyter Notebook,Excel,Statistics(A/B testing)

Database MySQL,Microsoft Azure

Tools Matplotlib,boxplot,Predictive model analysis using machine learning

Algorithms,Image classification using deep learning,sentiment

Analysis Using Natural Language Processing(NLP).

Projects

Movie recommender System- https://github.com/abhi19061991/movie-recommender ,


heroku deployment- https://mrs-dx.herokuapp.com/

PURPOSE-To recommend movies based on cosine similarity distances.


 Download pair of dataset from kaggle based on movie credits and movie name and later
merging them "on title".
Exploratory data analysis(EDA)
o data filtering using techniques such as drop columns,drop null values,remove
duplicated rows,checking the format
o The data to be stored in a “list” format followed by removing spaces and later merge
the required data columns into a paragraph and named it as "tags".
Data preprocessing and training
o We will vectorize the text using bags of words technique using hyperparameters such
as removing stopwords and max_features.
o Applying stemming and calculating the cosine distances amomg the vectors of the
same content.
o Sorting the first 5 movies based on the similarity index and print them.

Potato disease classification- https://github.com/abhi19061991/potato-disease-


classification
Purpose- Prediction of disease such as early and late blight in a potato plant based on visualization on
the surface of the leaves and classifying the images based on the 3 classes [early blight,late blight and
healthy

 Dataset from kaggle and imported in jupyter notebook with tf_dataset(tensorflow) in which
huge files is divided into classes and each element in dataset is a batch of 32 images.
 Splitting the data with 80% in training(conv2d) and rest 20% divided among validation and test
dataset.
 In preprocessing,rescaling and resize the image along with data augmentation .Trained the
image with filter of 32 and max pooling technique.
 90% accuracy in 5 epoch ,reduced loss function and good confidence score.

Sales Insights(power bi project)-


https://github.com/abhi19061991/power_bi-
Purpose - To unlock sales insights that are not visible before the sales team for decision
support and automate them to reduced manual time spent in data gathering.
 Firstly , we import the table to power bi tool from MySql and create the relationships(many to
one) in the data model section.
 In Power query editor,we tranform the data performed ETL(Extract,transform,load) and data
cleaning such as removing null values, converting sales in terms of indian rupee etc.
 Lastly ,in the report section we try to build a dashboard consisting of total sales amount,by
market etc.

Education
may 2011 - Jan 2014
Dayananda sagar college,Bengaluru - Bachelor in computer applications

You might also like