Archil Beridze
San Jose, CA • +1-669-***-**** • advtw2@r.postjobfree.com • LinkedIn • GitHub
Professional Experience
Data Scientist Intern May 2022 – August 2022
DeepCast, Houston, Texas
• Contributed to development of machine learning models reducing workload of scientists working in electrochemistry, some models could cut calculation time from 3 months to couple of minutes
• Collaborated with other data scientists to conduct state-of-the-art research and reconstruct architectures discussed in papers to identify baseline approaches,
• Assessed runtime of several open-source Python libraries and refactored code to optimize performance, resulting in 20% increase in runtime efficiency
• Utilized PyTorch framework to build neural network for solving curve fitting problem. Model converged 50% faster with trade-off in accuracy compared to classical statistical approaches Junior Data Scientist October 2019 – June 2021
Credo Bank, Tbilisi, Georgia
• Developed and managed credit scoring model. Scorecard was employed in Credo service centers on daily basis to determine creditworthiness of customers
• Utilized Pandas library to automate sales report generation. Report improved data accessibility for 70+ service centers of Credo throughout country
• Performed churn analysis of bank’s clients using classification and clustering algorithms
• Deployment of classifiers resulted in generating activity from 15% of inactive customers Education
Master of Science in Computer Engineering
San Jose State University, San Jose, CA May 2023
GPA: 3.86/4.00
Bachelor of Science in Computer Engineering
San Diego State University, San Diego, CA May 2020 GPA: 3.72/4.00
Relevant Coursework: Data Structures, OOP, Embedded Systems, Computer Organization, Machine Learning, Deep Learning, Data Mining, Intelligent Systems
Projects
Federated Recommendation System San Jose State University Master’s Project
• Leading small team of students to build an open-source federated recommendation system utilizing TensorFlow Federated library. Project aims to simplify usage of federated learning algorithms Smart Electrocardiogram System San Diego State University Capstone Project
• Designed neural network for processing and classifying electrocardiogram signals. Model was trained using Keras library and then implemented in C++ to allow it to be deployed on an FPGA device OXA programming language Personal Project
• Developed an interpreter for object-oriented, dynamically-typed, scripting language in Rust Awards
• Fulbright Scholarship 2021-2023
• Edmund S. Muskie Professional Fellowship Award 2022 Skills
Python, C++, Rust, Tensorflow, Keras, PyTorch, Scikit-learn, OpenCV, Data Science, Machine Learning, Deep Learning, Data Visualization, SQL, NLP, Computer Vision