Satwik Mishra

United States

@satwik_mishra191

Actively Interviewing 2021 SDE roles | MS CS@RIT | Ex-SDE Intern at Trimble Inc.

Badges

Problem Solving

Certifications

Work Experience

  • Software Engineer Intern

    Trimble•  August 2019 - December 2019

    ● Researched on ESRI's ArcGis GeoEvent Server and implemented a POC for automating and making the process of conversion of 811 Dig tickets to work-orders independent of original message format (JSON/XML/CSV) and sending the work-orders to Trimble Unity REST Endpoint. ● Integration of GainsightPX and Salesforce instance: fetch Trimble Unity KPIs like usage statistics, NPS ● Stress testing with NeoLoad and Datadog Synthetics. ● Tech stack: AWS EC2, DynamoDB, Python, JS, Node.js, Bash, Visual Scripting on Esri's GeoEvent Server, Apex, Visualforce Controllers, NeoLoad, Datadog.

  • Research Assistant

    Manipal Institute of Technology•  January 2018 - May 2018

    ● Developed an algorithm using 3D CNN with an ensemble of different activation functions and data augmentation to determine when lesions in the lungs are cancerous. ● Dataset: High-resolution CT Scans of lungs provided by Kaggle Data Science Bowl 2017. ● Libraries used - Tensorflow, Numpy, Pandas

  • Data Science Intern

    Apollo Munich Health Insurance•  May 2017 - July 2017

    # Fraudulent Claim Classifier ● Implemented a fraudulent health insurance claim prediction model. ● Understood the importance of data cleaning, processing, feature engineering and how to do it. ● Compared results by processing the data using multiple sampling techniques (SMOTE, Randomized Under-sampling), with proper cross-validation. ● Created an ensemble of Random Forest, Gradient Boosting and XGBoost as the prediction model. ● Analysed results using Confusion Matrix and ROC. ● Libraries used - Sklearn, imblearn, XGBoost, Seaborn # Sentiment Analysis on Twitter Feed ● Implemented a GUI to fetch tweets and give an analysis of the public sentiment for the input string passed. (eg. any company name) ● For NLP - tokenizing, removing stop words and POS- I used textblob which is built over NLTK. ● For computation of polarity, textblob uses a model trained on a Naive Bayes Classifier. ● Libraries used - Tweepy, TextBlob, NLTK, Wordcloud, PyQt

Education

  • Rochester Institute of Technology, Rochester

    Computer Science, MS•  August 2018 - Present

    CGPA: 3.81/4 || Courses:Mobile Robot Programming, Pattern Recognition, Intro to Big Data, Foundation of Intelligent Systems, Foundation of Computer Vision, Advanced Object Oriented Programming, Foundation of Computer Science Theory , Computational Problem Solving, Foundation of Algorithms

  • MIT, Manipal (Manipal Institute of Technology)

    Information Technology, B.Tech•  August 2014 - May 2018

    CGPA: 8.29/10 || Courses: ● Major in Information Technology : Data Structures, Designing Algorithms, Operating System, OOP, Principles of Digital Communication, Computer Networks ● Minor in Artificial Intelligence : Neural Network and Fuzzy Logic, Pattern Recognition, Computer Vision, Natural Computing and Human Computer Interaction.

Skills

satwik_mishra191 has not updated skills details yet.