Vishal Kumar-Chaudhary
Vishal Kumar-Chaudhary
With over 5.5 years of dedicated experience in end-to-end product development, I have consistently
demonstrated my proficiency in leading teams of more than 5 professionals to successfully deploy ser-
vices. My technical expertise spans Python, R, shiny, MySQL, mongodb, fastai, and PyTorch, and I
excel in full-stack development. I am adept at leveraging advanced database queries to solve complex
business problems efficiently. My strong foundation in statistics and mathematics further enhances
my problem-solving capabilities. Committed to delivering excellence, I bring a comprehensive skill
set and a proven track record to contribute effectively to my team and organization.
Corporeal Health Solutions as Data Scientist 1 Sep 2022 – April 2024 (1.5 year)
• Worked on invoice and financial statement OCR using pytesseract, paddleOCR, boto3 and the
using generative AI to improve the error if present which gave result almost 100% of the time
• Developed a large-scale NLP project with over 35K lines of code, where I contributed the majority
of the coding efforts alongside two developers. The project processed both Japanese and English
languages for text analysis.
• Working on shiny application with more than 6 inhouse R packages and 4-5 shiny module working
together, involving pdf processing, showing result in table, generating report using pagedown.
Also deployment via nginx as proxy-server with python API to give user ACL details.
• Working on disease detection from soft X-RAY images using pytorch which uses Deep Learning
to detect disease like TB, Normal, Covid etc.
• Leading team to improve product stability and
• Managing more than 10 repositories of the company, with 6 of them have heavy development
which I watch regularly and manage them at dev, review, admin level
Gyandata Pvt. Ltd. as Junior Data Scientist 2 May 2019 – Aug (3 years 4 months)
Project 1: Python Flask app development to deploy our NLP algorithm from scratch to deployed-
ment with the help of team of 4 members
• Managing and coding more than 60% of the codebase
• Full stack development using flask. Doing server management, login using auth0, nginx as proxy
server, redis server to share data between micro-services, shinyproxy to deploy shiny app, and
saving data into Mariadb server.
Project 2: Integration of payment and authentication for the a application and deployed the app.
Payment module is one micro service and shiny application was other micro-service which used auth0
for authentication.
• Integrated stripe as payment services and auth0 as my authentication service provider into my
web-application.
• Understanding of UI/UX with HTML, CSS , javascript, AJAX and backend technologies like
gunicorn to serve the flask application, jinja2 for dynamic HTML rendering, ORM for database
transactions.
• Created docker images for web application and shiny application. Run these images as services on
docker-swarm. Deployed MySQL-server outside and setup the connection between the application
to my dockerised MySQL server.
• Usage of open-source shiny deployment server i.e. shinyproxy which uses docker for scaling up
when the demand increases.
• Deploying on serving these micro-services and using NGINX as proxy server to connect to the
appropriate micro-service depending upon routing rules
Project 3: Keyword extraction from tweets to target and improve the particular product. Used
fasttext pre-trained as well as build new model with different hyper-parameters.
• Used n-gram as well as skipgrams to create word embeddings.
• Used k-means clustering to get the clusters with similar word
Project 4 : Demand forecast of Sales. Created interface for analysing and predicting trends and
extracting seasonality components automatically with the help of Fourier transform. Created Web
application out of it with core as the following machine learning models.
• Times Series Modelling: trend, seasonal, Non-deterministic decomposition(additive/multiplicative),
ACF, PACF, MA, AR, ARMA, ARIMA.
• Flask web application development in python using Flask, flask-sqlalchemy (ORM for database
operations), wtforms, flask-babelex.
Project 5 : Sensitivity analysis and recommendation for a data-set to increase a set of KPI in-
volving approx. 4000 variables.This project has some high variance and low variance variable, some
changeable and other non changeable so these variable cannot come in recommendation. Achieved
forward model accuracy 96.+% for 5 important KPIs. Recommendation and sensitivity found, were
sync in with domain knowledge.
• Understanding data, preprocessing, scaling, dimensionality reduction (PCA).
• Model building: Random Forest, SVM, Linear Regression, Logistic regression.
• Estimation: Hypothesis testing, confidence interval.
Coursework Electives:
Principles of Machine Learning , RL , Theory and Practice of Data Science , Proba-
bility, Stochastic process and Statistics, Blockchain Architecture and its use cases,
Design and analysis of algorithm, Information Theory and Coding, Linear Algebra and Numerical
Analysis, Graph Theory
Hobbies Playing Guitar, Chess, reading books.