Nirav Patel

Raleigh, NC · (973) 420-0506 · nkpatel8@ncsu.edu

Graduate student in Computer Science at North Carolina State University (Graduating Summer 2022).
Previously Software Engineer at Juniper Networks .

Experience

Software Engineering Intern

North Carolina State University
  • Designed graph schema for the GHCN-D climate dataset and automated building Neo4j graph database in aws. Benchmarked performance, scaling and cost of maintaining the database.
  • Designed Schema for GHCN-D dataset for AWS DynamoDB. Created AWS cloudformation stack consisting s3, lambda and sqs to build database in dynamoDB directly from csv file. Compared the benchmarking results of DynamoDB to Neo4j.
  • Created GraphQL API using Grandstack and HTTP API to access Neo4j database.
  • Performed exploratory data analysis of the database in python and different visualization in tableau.
May 2021 - Present

Software Engineer 2

Juniper Networks
  • Worked on enhancing multicast snooping functionality over multiple platforms - routers and switches running JUNOS operating system. Worked closely with Juniper customers to resolve various issues concerned with memory management, scheduling, threading, snooping behaviour in network running different protocols such as EVPN, H-VPLS,EVPN, pim, igmp, mld,bgp
  • Worked on makefile for migrating multicast snooping process from JUNOS(BSD based Juniper OS) to EVO(Linux based Juniper OS). Resolved various cyclic dependencies and undefined symbols as the code span over million lines of code.
  • Developed a feature to read traffic statistics from PFE and showing traffic summary in UI at RE level.
June 2018 - December 2020

Software Engineer Intern

Juniper Networks
  • Developed robust test scripts verifying BUM traffic for multicast snooping daemon running evpn on QFX series platform using Spirent and robot framework
  • Built different network topologies and automated testing of multicast snooping in different types of routing instance running protocols such as ospf, bgp, igmp, pim, p2mp, vpls
January 2018 - May 2018

Research intern

DAIICT
  • Manageed to get 93% accuracy to mark the attendance using an image of a classroom. Used Haarcascade classifier and adaboost to detect the faces and CNN to recognize the face
May 2017 - August 2017

Education

North Carolina State University

Masters of Computer Science
Coursework - Automated Learning and Data Analysis, Neural Networks and Deep Learning, Database Management System, Graph Data Mining, Software Engineering, Design and Analysis of Algorithms

GPA: 4.0/4.0

January 2021 - Present (Expected Gradudation: May 2022)

Dhirubhai Ambani Institute of Information & Communication Technology

Bachelors of Technology in Computer Science
Coursework - Neural Networks, Computer vision, Information retrieval , Computational data science and analytic, Data mining and warehousing, Internet of things

GPA: 7.6/10.0

August 2014 - May 2018

Skills

Programming Languages & Frameworks
  • Java
  • Python
  • C
  • SQL
  • Cypher
Tools and Technologies
  • AWS
  • Docker
  • Neo4j
  • Tableau
  • Robot Framework
  • Git
  • Spirent
  • Spring Boot
Libraries
  • Keras
  • Tensorflow
  • Pandas
  • Pytorch
  • Numpy
  • Sklearn
  • OpenCV
  • Seaborn

Interests

I love playing sports. I am a good chess, badminton and cricket player. On weekends I love binge watching TV shows. And I am always up for hiking.

I consider myself enthusiastic data scientist and I spend good amount of time learning concepts and reading new algorithms/technologies coming out in the constantly changing world.

Projects

Google Landmark Recognition

Built a deep learning model to recognize a landmark given an image. Used pretrained resnet101 as a backbone architecture and used transfer learning to train fully connected neural network. Compared results of this model with different loss functions such as cross entropy loss, arcface and triplet loss.

Big Box Retailer Database Service

Designed and implemented a database service for a Costco like store using MariaDB & Spring Boot. Performed partitioning, indexing, and query optimizations to improve performance

Summarization of tweets

Developed a model to give a summary of people's view on twitter given a hashtag. Created the dataset by randmoly selecting trending hashtags using twitter APIs. Used text mining methods to clean the dataset. Implemented modified tf-idf algorithm and K means to get the summary.

Human Activity Recognition

Given a time series data of accelerometer and gyroscope values, implemented different models to predict correct activity label. Compared results of classical models such as SVM, XGBoost, bagging classifier, LSTM, Bidirectional LSTM and CNN-LSTM after extracting handmade features. Achieved a F1-score of 92\% on a highly imbalance test set.

Protein sequence prediction

Used Machine Learning algorithms to predict the type of protein. Implemented and compared different neural network \newline classification methods MLP, RBF, MRAN in MATLAB

Smart traffic signal

Built a traffic signal system to compute the timer value of the signal based on the number of vehicles on each side of the signal. Used edge detection and background subtraction algorithm for counting the number %of vehicles from the image taken by raspberry pie.

Contact

If you liked my work or have queries about any of my projects, you can reach out to me via email or DM me on Twitter. If you'd like to contribute to any of my projects you can simply open an issue on GitHub.

nirav1929
nkpatel8@ncsu.edu