Nimit Johri add8ed@r.postjobfree.com

**** * *** ****** ** Philadelphia PA 19122 +1-267-***-**** EDUCATION:

Temple University, College of Science & Technology, Philadelphia, PA Master of Science in Computer Science, Graduation: May 2020 GPA: 3.47 Dr. A. P. J. Abdul Kalam Technical University, Lucknow, U,P. India Bachelor of Technology – Information Technology, Graduation: May 2016 TECHNICAL SKILLS

Concepts: Data Structures, Algorithms, Advanced Database, Graph Theory, Operating Systems, Cloud Map Reduce

Languages: C#, Java, PHP, SQL, Ajax, HTML, CSS, JavaScript, Apache Spark PROFESSIONAL EXPERIENCE:

Software Engineer Temple University, Philadelphia PA September 2018 – present

• Built a web application using PHP which stores research studies provided by users & metadata fetched with an API and stores it on Github as cloud storage, reducing the team workflow by 20%

• Designed a Fuzzy Matching algorithm in PHP which implements Levenshtein distance & database pipelining. Implemented database indexes to use index only queries reducing data fetch times by 77%

• Implemented master slave architecture of parallel processing for the fuzzy matching algorithm which terminates within 0.08 times of the select query execution time reducing application runtime by 94%

• Made the solution scalable by implementing Ajax throttling to overcome modern browsers limit of scheduling around only 2700 parallel Ajax requests at a time which prevents DOS attacks. Assistant System Engineer TATA Consultancy Services, India July 2016 – June 2018

• Programmed a POC in C# to automate multiple piped workflows reducing workflow time by 33%

• Automated monthly logging of multiple applications in a single yearly excel file using .NET framework for the client, supporting the piped workflow and reducing halts in the workflow by 75%

• Built SQL scripts to combine and fetch data by joining tables from different production databases together as per the requests from client side business analysts, reducing the team workflow by 25% Intern Air India, India Summer 2015

• Rationalized the business requirement of implementing wireless network over an existing wired network

• Prepared a report suggesting a Wi-Fi router layout in Air India aircraft maintenance hangar with one less router node reducing the implementation costs by 16% from the original plan PROJECTS:

• Divorce prediction & marriage counseling with Neural Network – Trained a neural net on a divorce dataset based on a Likert scale survey with 98% accuracy. Learned core causes with feature selection and built an algorithm to suggest improving on correlated causes as counselling if divorce is predicted.

• Friend Suggestion algorithm – Developed an algorithm to suggest friends by implementing the principle of triadic closure in an undirected and unweighted social graph. Implemented a ranking system for the nodes suggested as friends based on the potential of number of social triangles closed

• Self-learning BlackJack algorithm with Q-Learning – Created an AI using Q-Learning, a type of Reinforced learning. Implemented a system to reward it for good moves and penalize on bad moves and trained it by letting it play against itself. The trained model could make about 80% correct moves.

• Comparing performance of Relational, JSON and Hybrid database models – Analyzed the query execution plans for same queries on each model to compare cost of operations and optimize them. Finally concluded that Relational was fastest and JSON was 300% slower while hybrid was only 100% slower.

• Linux mini-Kernel as a web service – Developed a Linux kernel to parse and execute commands, pipes and redirection. Implemented distributed tuple space to receive command and respond on return socket.