Kevin Kochpatcharin Email & Phone Number

San Ramon, Ca, Us

West Lafayette, In, Us

Research in Applications for Learning Machines (REALM) Consortium is an industry-academia partnership between Northrop Grumman Corporation and multiple universities. I am in the team of Situational Knowledge On Demand (SKOD), which is headed by Prof. Bharat Bhargava (Purdue), Prof. Michael Stonebraker (MIT), and Prof. Michael Cafarella (U of Michigan). It is a multi-year project to harness relevant information from heterogenous data sources (tweets, traffic cameras, dash cams, body cams) for agents in the field when needed. The team consists of three Professors, and eight PhD students. My job responsibilities: • Trained computer vision algorithms to detect pedestrians from street camera footage and pedestrian attributes to allow for future reidentification.• Ran experiments to benchmark object detection and pedestrian attribute recognition models on street camera and dashcam• Surveyed research papers and contributed new ideas to the team such as the reidentification of pedestrians by extracting recognizable attributes.• Improved method to scrape relevant text data from Twitter and build a database of hundreds of thousands of tweets.• Used Python, Pytorch, and AlexeyAB implementation of darknet (convolutional neural network used for feature extraction from images).• Published Paper: Surveillance Video Querying With A Human-in-the-Loop at HILDA 2020 workshop (Human-In-the-Loop Data Analytics)

West Lafayette, In, Us

Collaboration between Prof. Bharat Bhargava (Purdue) and Chris Jenkins (Sandia National Lab). Research focus on application of Moving Target Defense methods (MTD) on MIL-STD 1533 (serial bus communication standards). MTD randomly changes system attributes to greatly increase the effort required for an attacker. My job responsibility is to defeat the MTD protocols using machine learning.• Experimented with machine learning attacks on the two MTD protocol implementations from Sandia. I was given message histories generated by Sandia. I then used machine learning to determine the vulnerabilities in MTD that can be learned by the attacker. • Identified components of ESA SpaceWire standard (similar to MIL-STD 1533) that could be improved by application of MTD.

West Lafayette, In, Us

I became a TA since junior year BS, for a total of four semesters. I teach upper division CS classes. My job responsibilities:• Cloud Computing: This is the second time this class is offered, by Professor Douglas Corner. As the class is still in development, my job includes working with professor to design lesson plan. I also create and grade lab assignment. I write homework answer keys. The assignments are designed to teach students to both employ Docker Containers and Kubernetes; and to deploy services on Amazon AWS, Microsoft Azure, and Google Cloud Platform. • Operating System: I oversaw lab sessions where students implement features such as scheduler, interrupt handler, system calls in XINU OS in C++ and X86, graded homework and reviewed exams• Relational Databases: I was TA for this class for two semesters. I taught ACID requirements, Spark, Hadoop, MapReduce. I oversaw lab sessions and reviewed and graded homework and exams

West Lafayette, In, Us

Research to investigate blockchain to augment the WaxedPrune information sharing technology to share trusted private health information between untrusted parties. • Combined blockchain and extended attribute aware active bundle technology to improve data sharing privacy, authorization, and leak detection.• Designed and executed performance testing for the blockchain network.• Measured transaction latency from recording data sharing operations with IBM Hyperledger.• This project won 2nd Place Poster at CERIAS (Center for Education and Research in Information Assurance and Security) Symposium 2019.

San Ramon, Ca, Us

Five9 is a leading provider of cloud contact center software, headquartered in San Ramon, CA. I was hired during the pandemic remotely and worked remotely. I worked with a team of eight developers (located in Russia and San Ramon, CA). My job responsibilities:• Developed a set of tools to verify the correctness of call events between on-premise Apache Kafka clusters and GCP datacenters. The tools monitor dataflow, latency, and replication of call events streaming within various internal upstream and downstream systems• Implemented tool to detect issues in Kafka Mirrormaker that were preventing correct replication of customer dataflow.• Used Java, Gradle, Apache Kafka, Google Cloud Storage, GCP PubSub• Assisted the other intern with SQL and Python scripts for his tasks

Master Of Science - Ms, Computer Science

Bachelor Of Applied Science (B.A.Sc.), Computer Science

High School Diploma, General Education