ac7kww@r.postjobfree.com HARSHA KERAGODU SHIVAPPA 312-***-**** Chicago Illinois

https://www.linkedin.com/in/harsha-keragodu-shivappa-8a16a99b/ github.com/harshaks23 devpost.com/HarshaKeragoduShivappa EDUCATION:

Illinois Institute of Technology, Master of Science in Computer Science, Chicago, IL. (GPA 3.6/4) Graduating in Dec 2018 Coursework: Data Mining, Algorithms, Machine learning, Cloud computing, NLP, AI, Game Theory, Big data Technologies, Android, computer networks VISVESVARAYA INSTITUTE OF TECHNOLOGY, Bachelor of Engineering in Electrical Engineering, Mysore, IN. (GPA 3.8/4) July 2016 SKILLS

Languages: Java, Python, C++, R, PL/SQL, Shell BigData : Spark, Hadoop, Apache Hive, Pig, Pyspark,HDFS Tools: Android Studio, MATLAB, Android sdk, git Databases: MongoDB, MySQL, DynamoDB, PostgreSQL, Firebase, SQLite, SQL Alchemy Web : JavaScript, HTML5, CSS, Django, Flask, Nginx, Heroku ML: TensorFlow, Keras, SpaCy, pandas, numpy,sckit-learn, Theano, OpenCV, Gensim,Polyglot, NLTK, CoreNLP, matplotlib, TextBlob, RASA NLU / Core, NLP Orange, Weka, Tableau,Shiny, Jupyter, Pycharm, sklearn, BeautifulSoup . CLOUD: Amazon AWS, Alexa, Lex, Lambda OTHER: Docker, Json, XML. PROFESSIONAL EXPERIENCE:

SOMOS Polsky Center for Entrepreneurship and Innovation,University of Chicago June 2018-present

• SOMOS is a bootstrapped nonprofit startup by students that provides immigrants a secure online community. Currently building a prototype to be launched in Chicago.

• Building a full-fledged Django web-app with Postgres, nginx and OOP design. DATA MINING TEACHING ASSISTANT Aug 2018-present

Involve in teaching, grading and conducting quiz to graduate students and reinforce learning concepts involve about data mining, machine learning and natural language processing to the graduate and undergrad students: Python, R, Weka, Orange and TensorFlow DIGITAL INTENT, Chicago Data Science Intern Sep2017-May2018

• Build time series models in predicting the deliveries of the food on the EAT-Purely data (food delivery company like Grubhub).

• Reviewed and documented existing application code, involved in code clean up, optimization and versioning.

• Optimized restaurant delivery boundaries to maximize delivery driver efficiency and predicted future order volume to plan for delivery driver staffing needs. Provide most up-to-date tracking of loss trend and change.

• Worked on NLP algorithms, clustering, sentimental analysis and classification of the reviews and invoices.

• Analyze consumer feedback and build models to identify and prioritize consumer feedback. Assisted with migrating the psql database to aws

• Assisted in Developing apps in Android (Eat Purely Team). PICKING ME FOUNDATION, Chicago Web Developer Intern July 2017 -Aug 2017 Developed the website using fire spring framework(www.pickingme.org) as an intern for the nonprofit Foundation, currently responsible for writing business plan, product development, fund raising, supply chain management, market research and hiring. INDIAN INSTITUTE OF SCIENCE (IISC), Bangalore Research Scientist Oct 2015- May2016 Worked on Nano technology, utilized machine learning and NLP algorithms to do classification, labelling, relevance matching, prediction on weather data collected by drones, weather monitors.

ACADEMIC PROJECTS

CHAT BOT: Built a dynamic chat bot that chats with the user particularly for website like reedit, the framework can be used for other platform by processing the different data (AP news Word2Vec, LSTM/ Recurrent Neural Network, Deep Learning, keras,theano,NLP,Genesim,flask.). SAFETY VISION: Real-time face and object detection using OpenCV, this helps in detecting people along with objects and then give notification to the through Alexa and SMS (Haar cascade for face detection, LBPH for training and detecting the faces, OpenCV, Cascade Classifier) TENSORFLOW: Built various deep learning models (CNN, RNN) in image detection, text classification and sentimental analysis, Feature Engineering, transfer Learning, autoencoders using unsupervised learning and other predictive models using tensor flow DETECTION OF CRIME RATE: Built various models (Decision tree, Gaussian NB, Linear and nonlinear SVM, K-nearest neighbor, CV, Random forest, Ridge and linear regression, Neural MLP) to predict the crime rate based on demographic and economic info. FRAMEWORK FOR THE ANALYSIS OF BIG HISTORIC TEXT COLLECTION: Build classification and clustering models to disambiguate all mention of “Chicago School” and used Ensemble learning to improve the performance. preprocess 1 TB of Architectural data (100k Books) Active Learning: Implemented learning with rationales technique by manually providing rationales for 300 training instances. Compared learning with rationales and without the rationales. Implemented the active learning framework. Chicago Assistant Chi-Hack Braintree, Chicago

Building a Chicago assistant/chat bot, a prototype to be submitted to the Chicago mayor office: a text (Messenger), voice interface (Alexa and Google Assistant) for requesting city services (311 Service Requests) and making inquiries about city services. Tech: RASA NLU/ CORE. Docker, Alexa, Google assistant Hospital Admissions: Built various predictive models in predicting hospital readmissions for diabetes patients on a health data using R (SVM, naïve Bayes, decision tree)

Tera Sort benchmarking: Developed the benchmark to measure bandwidth and latency of network and calculate the IOPS and FLOPS for CPU. Developed benchmark for sorting 1TB data on different node cluster on amazon aws EC2 using techniques, shared memory, Hadoop and spark (JAVA and python) Link-State Routing Simulator Simulated the process of generating forward table for each router in each network and computed optimal path with least cost between any two specific routers. Used Dijkstra's algorithm for finding shortest path. Design and development of mobile application: Sticky-Notepad, Stock Watcher, Food selling and News Gateway app on Android. Hacks at University of Michigan Ann Arbor, Purdue, University of Chicago Urbana Champaign and University of Washington.

• ARDrone 2.0 Implemented framework in python to automate the drone and detect object (more than 10) and face recognition on both real time video streaming and image snaps was monitored using HERE maps .Implemented also on Mobile and webcam so that people in danger during floods, disaster can use (Mobile Nets, Single shot detectors, OpenCV,deep learning, Twilio, Clarify, NodeJS, Webpack, Google cognitive and Microsoft api, SocketIO ).

• Distributed Credit Bureau using Blockchain. Created a blockchain prototype and built API using flask and applied Predictive, time series modeling and real time analysis using Microsoft Cognitive services.

• Speakly and DSDP Diagnosing Childhood Speech and Language Development Disorders disorder by using examples from a doctor. Sleep Deprivation and prevention using Facebook data. Analysis and Prediction using predictive models Alexa, AWS lambda, flask, chart.js, NLP, MongoDB, Docker.