I am Devashish, a software engineer
and a data science enthusiast.
Python
Java
JavaScript
Backend
Frontend
Databases
Data Science
Cloud Technologies
About Me
Currently, I am pursuing my masters in Software Engineering at San Jose State University.
Here I am specializing in Data Sciences.
I have done several projects that included designing UI wire-frames, creating back-ends, designing
databases, creating front-ends, creating data analytical dashboards, predicting insights and hosting the
web applications.
I have also helped some organizations in creating their data engines. I am always looking for new and
challenging tasks. My goal is to make valuable contribution in all the fields I am working on. I am also
a past-time soccer player and a self learned musician.
The project is aimed to create a complete data engine for Swastika. I am currently working on this. I am doing
this just for my learning.
Design databases to store client transactions, stock prices, employee data, etc.
Write ETL scripts to store data.
Analyze customer data by extracting features and cluster customers on the basis of their
value.
Predict when a customer is going to make the next transaction and what amount is he/she
going to spend.
Create a full-stack application to help Swastika watch their customer behavior.
Enterprise Backend as a Service
The project is aimed to build an full-stack application that will interact with the user to
create a backend
code that will create the required product with bare minimum requirements starting from the
creation of the database to the creation of the APIs to communicate with the database. In
addition
to this, we will provide a feature that will predict useful insights based on the data added
through
our APIs using Machine Learning services.
This project is my final year masters project which I am working on with a team of 4.
Work Experience
Enovix Corporation
I worked at Enovix as a software engineer intern from May 2019 to October 2019.
Built 19 web applications from scratch to migrate traditional VBA excel uploaders to
improve performance benchmarks. These tools were used across the organization by all the
engineers and operators.
Used regression machine learning algorithms to predict the life of a battery and outlier
detection algorithms to detect defects in batteries. This saved waiting for 600 life
cycles to finish.
Designed databases with ETL and built data visualization and reporting tools to support
test engineering, R&D.
Hosted these tools on Microsoft IIS to increase availability.
Technologies: Git, Django, Python, REST, Microsoft IIS, Vue.js, Handsontable, MS SQL,
MySQL, and Sklearn.
Mediaocean
I worked at Mediaocean Pune as an associate software engineer from July 2017 to July 2018.
Wrote 113 test cases and assisted in data transformation, quality, profiling,
validation, and reconciliation.
Build a tool in MOHack to track the user flow using google tag manager. This tool helped
product managers to see how the users were actually using the product.
Collaborated on the automation framework and increased the quality of the product
(PRISMA).
I worked at MiniOrange as a software engineer intern from Jan 2016 to June 2016.
Built Drupal OTP Verification Module. This plugin has 3000+ active installs.
Collaborated on miniOrange’s Social Sharing plugin for WordPress. This plugin has 20000+
active installs.
Supported in Social Login Analytics and search engine optimization.
Project Work
Medium Stories - Analysis and Predictions
The project was aimed to help the user write posts on Medium. The program assisted by
predicting the next word and insights like tags, reading time, number of claps the
article is going to get.
NLP, Recursive Neural Network, Multivariate Regression, Random Forest, Naïve Bayes were
used.
A short 3 minute video to describe how information retrieval on music works.
ML on Source Code
A short 5 minute video explanation about machine learning on source code.
Tanks A Virtual War
This project was aimed to create a desktop tanks multiplayer game.
Networking principles like UDP protocol, TCP protocol etc were used in this project.
Design patterns like observer pattern, mediator pattern, component pattern etc were used
in this project.
Movie Recommendation System
Project aimed to create a movie recommendation system using Spark. Different algorithms for
collaborative filtering
like item-item collaborative filtering, SVD and ALS were used.
Stack overflow - Analysis and Predictions
Project aimed to utilize stack overflow annual survey results and predict users salary based on
other features.
Spartan Messenger App
Project aimed to develope a Spartan Messenger using GRPC in Python3. Used gRPC response streaming
to continuously receive chat messages from the Spartan server.
Traffic Image Classification for Traffic Lights
Project aimed to classify images into 15 different types of vehicles by extracting features from
images like Histogram of
Oriented Gradients features, Normalized Color Histogram features, Local Binary Pattern features,
Color gradient
features, Depth of Field features, etc. and then classify them.
Chicago Crime Visualization Dashboard
Project aimed to create a dashboard to visualize various aspects of Chicago crime data.