Harsha Priyatham Yarravarapu

+1-602-***-**** ad14f3@r.postjobfree.com github.com/hy103 https://linkedin.com/in/harsha-yarravarapu EDUCATION

Master of Computer Science, Arizona State University, Tempe, AZ Jan 2023 – Dec 2024 Courses: Foundation of Algorithms, Statistical Machine Learning, Image Analytics & Informatics, Engineering Blockchain Applications, Introduction to Digital Image Processing GPA: 4.0/4.0 Bachelor of Technology, Indian Institute of Technology Bhubaneswar, India Aug 2014 - Jun 2018 Courses: Programming and Data Structures, Computer Networks, Operating Systems, Database Systems GPA: 7.55/10 TECHNICAL SKILLS

Languages and Databases Python, R, C++, Java, C, HTML, CSS, Oracle, MySQL, Snowflake, BigQuery Libraries and Frameworks Scikit Learn, PyTorch, TensorFlow, Keras, PySpark, AWS, GCP Location Tools Anaconda, IntelliJ, XCode, GitLab, Jira, CLM, Postman, Apache Airflow, Jenkins Certifications Coursera certified SQL Developer, Generative AI by Google WORK EXPERIENCE

Machine Learning Engineer ~ SDE II - Talentica, Pune, India Oct 2020 - Dec 2022 Time Series forecasting for TV show ratings - WIDEORBIT Technologies: Python, R, Snowflake, Apache Airflow, AWS (EC2, sagemaker,S3,Redshift), GCP, Tableau

● Built a cloud-native application offering television channels an advanced rating service. Leveraged strategic ad slot placement to attract investments from 3 additional channels.

● Created an application for daily monitoring of actual viewership and predictions. Automated the process by developing log-generating scripts saving 40 man-days of developer effort per year

● Achieved 15% MAPE, 83% accuracy, and 87% reliability for prime TV timing slots which are the best in the industry Real estate appreciation - Mynd

Technologies: Python, R, Java, PyTorch, Snowflake, Apache Airflow, PySpark, Jenkins, AWS (Redshift, S3)

● Constructed ML pipeline and implemented optimal bid range facility for property quotes on AWS

● Designed ML pipeline to handle batches of incoming data and deployed ratings API

● Scaled the ML pipeline from handling 1,000 requests per day to supporting 5,000 requests per day to be submitted in a single workload by identifying and solving bottlenecks Software Engineer – HCL, Chennai, India Oct 2018 - Sep 2020 Concall-Meeting Notes

Technologies: Python, PyTorch, Tensorflow, Spark, Postman, GCP, BigQuery

● Designed Speech recognition system using Bi-LSTM models for conference calls

● Increased speech-to-text conversion rate to 96% by devising grammar and customizing, and training the acoustic model

● Reduced the model training time by 42hrs by implementing the parallel processing on cloud clusters E-commerce product recommendation

Technologies: Python, R, Java, Python, R, Tableau, Java script, HTML, CSS

● Developed a recommendation engine designed to offer personalized product suggestions based on user preferences.

● Implemented real-time recommendations by taking user cart and watchlist data and created batch recommendations.

● Enhanced user engagement by 4% through collaborative filtering, displaying similar items based on user preferences ACADEMIC PROJECTS

● Petrinet for Hyper Ledger Fabric (Python, C++, SPNP) (Research Assistant at ASU)– Designed PetriNet model to substitute the Hyper Ledger Fabric and set up micro-benchmarks for performance analysis of PetriNet model.

● Posture Corrector (Python, MediaPipe, TensorFlow) - Implemented Pose Landmarker to correct the postures of cricketers using the MediaPipe library and created the batting shots posture configuration files.

● Recommender System (Python, Spark) - Modeled an intelligent musician recommendation system based on the user’s listening history using a collaborative filtering technique and alternating least squares (ALS) algorithm

● Database Systems (Java, JDBC, MySQL) - Designed a scalable database for a Hotel Management System inMariaDB which performs end-to-end functionalities of customers and employees with GUI developed in Java ACTIVITIES AND ACHIEVEMENTS

● Awarded Go Getters award in Talentica for successful migration of Real estate appreciation from AWS to Snowflake

● Received Best Employee Award in 2019 in HCl for deployment of quantitative assessment of credit fraud detection.

● Teaching Assistant, Principles of Programming in C++, mentoring a class of 190 students at Arizona State University

Contact this candidate