HIMANSHU LONDHE

410-***-**** adlpw0@r.postjobfree.com United States

github.com/himanshulondhe Website: himanshu-londhe.me linkedin.com/in/himanshu-londhe EDUCATION

University of Maryland, Baltimore County, USA May 2021 Master in Computer Science [ GPA: 3.8 ]

University of Pune, India June 2018

Bachelor of Engineering, Computer Science [ GPA: 3.7 ] SKILLS

Languages Python, C++, C, R, Java, HTML, CSS

Librarires NumPy, SKlearn, Apache Spark, TensorFlow, Keras, Pandas, Matplotlib, React Web Technologies RESTful APIs, Git, Node JS, Flask, Django Database MySQL, SQLite, MongoDB, Cassandra

Other Linux/Unix Administration, Shell Scripting, AWS, Debugging, Arduino, Agile, Android, MPI, Network Technology, Android, ML-random forest, regression, decision trees, SVM, Naive Bayes WORK EXPERIENCE

Software Engineer Intern j Ardent Privacy: Baltimore, MD July 2020 - Sept 2020

Developed and implemented a novel machine learning framework from scratch to identify the presence of sensitive data so as to nd indicators for the data minimization platform without scanning the content of the les, thereby preserving user data privacy.

Built RESTful APIs using python and ask.

Designed a SQLite database schema to store client server metadata of upto 50GB. Used indexing techniques to optimize database. Utilized Redis for fast access to API responses and data caching.

Maintained code and work ow e ciently on BitBucket. Responsible for the Jira admin role for the project.

Integrated and deployed the model successfully on Amazon EC2 web server. Deployed using AWS CI/CD tools like AWS CodeCommit, AWS CodeBuild, AWS CodeDeploy, and AWS CodePipeline.

Monitored API health using Grafana to ensure high availability of the service.

[ Python, Flask, AWS, SQLite, sklearn, Back-end ]

RESEARCH PUBLICATION June 2018 - Dec 2018

Paper on Enhanced Support Vector Machine with Speed Up and Reduced Sensitivity accepted and published in International Journal for Research in Applied Science and Engineering Technology (IJRASET) Volume 6 Issue XII, Dec 2018- Available at www.ijraset.com

PROJECTS

Enhanced Support Vector Machine with Speed Up and Reduced Sensitivity Aug 2017 - Dec 2018 Improved the classi cation accuracy of linear Support Vector Machines by 8-13% by designing a data prepossessing module which reduces ’scatteredness’ of the data. [ Python, sklearn, pandas, Matplotlib ] Explainable AI for Air Quality Prediction as a Full-Stack Application Sept 2020 - Dec 2020

-Designed and developed a classi er for calculating Air Quality Index from the weather data with 98% accuracy and using Explainable AI to explain the results of the Classi cation Model.

-Took ownership of designing RESTful APIs with python and Django.

-Implemented front-end service using React, javascript and HTML CSS.

[ Python, XGBoost, Regression, Multi-label, full-stack, Classi cation, Django, React JS ] Centralized Multi-User Concurrent Bank Account Manager Sept 2019 - Dec 2019

-Designed and developed a bank server that handles multiple clients and does so using distributed programming concepts.

-Devised socket programming and TCP/IP protocols to handle concurrent transaction requests.

[ C++, sockets, mutexes, fault tolerance, time synchronization ] Fall Detection Classi er using CNN - Computer Vision Feb 2020 - May 2020

-Implemented a Concurrent Neural Network to process and classify sequences of video streams into fall/no-fall events with an accuracy of 92-95%.

-Used the dense ow tool to isolate movements in a videp sequence of input RGB images.

[ Python, SKlearn, keras, tensor ow ]

CERTIFICATIONS

R 101, (RP0101EN, provided by Cognitive Class) an online course on cognitiveclass.ai C++, IIT Bombay (Spoken Tutorial).

JAVA, IIT Bombay (Spoken Tutorial).

Contact this candidate