Pratik Karnik
-
- Mumbai, IN
www.github.com/pratikskarnik
https://www.linkedin.com/in/pratikskarnik/
https://www.kaggle.com/pratikskarnik
Machine Learning Engineer
SUMMARY
2 years 9 months of experience as a Consultant in L&T InfoTech. Implemented Machine Learning and Deep Learning projects using
Python. Looking for a Data Science role for personal and organizational growth.
SKILLS
Machine Learning, Deep Learning, Linear Regression, Logistic Regression, Statistical Modelling, Decision Trees, Predictive
Modelling, Ensemble techniques, Boosting, Clustering, Image Classification, Computer Vision, NLP, Data Cleaning, Data
Analysis, Data Visualization, Model Evaluation and Deployment, Python, Scikit-learn, Tensorflow, Keras, NLTK, Spacy,
PyTorch, FastAI, Rasa, Java, SQL, MS Excel, HTML, CSS, JavaScript, Docker, Azure, Heroku, Android Studio, Flutter
ACADEMIC PROJECTS
Name: Indian Food Image Classifier | Tech Stack: Python | June '21
Objective: Based on the given image identify the type of Indian food.
Solution: Designed Resnet50 Image classifier using Deep learning, Fast.ai and Python to predict the type of food from 250+
varieties of food. Deployed the trained deep learning model on Heroku using HTML, CSS and JavaScript. Created an Android
app with Flutter which uses the deployed deep learning image classifier model to classify 250+ varieties of Indian Food.
Key Achievement: Predicted the type of food with an accuracy of 56%
Link to Deployed web app: https://indianfoodclassifier.herokuapp.com/
Name: Entity Recognition in Healthcare | Tech Stack: Python | July'21
Objective: Based on the given text file extract diseases and their corresponding treatments.
Solution: Extracted PoS tags out of labelled text data. Used NLP techniques to extract features out of words and sentences.
Designed CRF model using NLP, Python and Spacy to extract diseases and their corresponding treatments from the text.
Key Achievement: Found Diseases and corresponding treatments by developing a model with an f1 score of 90%.
Name: Amazon ML Engineer Hiring Challenge | Tech Stack: Python | Nov'20
Objective: Based on the given data for an ecommerce website, identify whether to target the person for leads or not
Solution: Cleaned the data by filling the NaN values and removing the outliers. Used a recursive feature elimination technique
based on p-values. Developed Machine Learning classification models using Logistic Regression, Random Forest, XGBoost and
SVM and selected the best Machine Learning model. Explained the entire code in a medium article.
Key Achievement: Got a Precision score of 93.7% on test data.
Link to code: https://github.com/pratikskarnik/Amazon_ML_Engineer_Hiring_Challenge_2020
Link to article: https://becominghuman.ai/how-to-land-an-amazon-ml-engineer-interview-and-mistakes-to-avoid-1f5b87d22702
PROFESSIONAL EXPERIENCE
Aug '18 - May '21
Consultant - Package Implementation
Mumbai, IN
L&T InfoTech
Security Administrator
Administrated security requests for Johnson and Johnson client in JDE ERP for 2 years.
Handled all security requests related to major projects independently in a client facing role.
Software Engineer
Created a Rasa Chatbot to track new users registering on JIRA.
Created simple web app using Angular to display pivot table.
EDUCATION
Post Graduate Diploma in Data Science
IIIT Bangalore & upGrad
Oct '20 - Nov '21
Bengaluru, IN
Current GPA: 3.71/4.0
Bachelor of Engineering in Computer Engineering
Vidyalankar Institute of Technology
Secured 7.09/10.0
Aug '14 - Jul '18
Mumbai, IN
CERTIFICATIONS
DeepLearning.AI Tensorflow Developer Specialization Coursera
Data Science Nano Degree Imarticus
Neural Networks and Deep Learning Coursera
Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization Coursera
Structuring Machine Learning Projects Coursera
Natural Language Processing Kaggle
Facial Expression Classification Using Residual Neural Nets Coursera
Classification with Transfer Learning in Keras Coursera
Python for Data Science and AI Coursera
Databases and SQL for Data Science Coursera
ADDITIONAL INFORMATION
Qualified for Google Code Jam Round 1 with rank of 7844
Languages: English , Hindi , Marathi
