Chiharu Akiyama
*******************@*****.*** ***- 660-7543 Portland, OR 97206
Objective
Recent Computer Science graduate seeking junior software engineer position to apply my skills while learning to develop high quality applications.
Education
2019 - 2022
Portland State University, Portland, OR
Bachelor of Science, Major in Computer Science
GPA 3.70
Experience
• SightLine Applications, Inc
Software Engineer - Senior Capstone Project (June 2021 – Dec 2021)
- Tuned image classifier for Oregon-based onboard video processing company, SightLine Applications, Inc to achieve accuracy of 95% or more.
- Collected, cleaned, and preprocessed images for the custom dataset.
- Trained and tested SightLine Application’s proprietary image classifier on the custom dataset.
- Participated in weekly team meetings for progress updates and brainstorming.
- Identified that some of the classifier accuracy issues were due to the classifier learning background textures instead of the object in images. Technical Skills
• Languages: C++, JavaScript, Python,
HTML, CSS, SQL
• Frameworks: React
• Libraries: Tensorflow, Pytorch, Keras,
Numpy, Pandas
• Version Control: Git/Github
Additional Skills
• Team-player
• Time management
• Self-motivated
Relevant Coursework
• CS 410P Front End Web Development
- Created a beginner-friendly dashboard application that displayed and compared user- selected cryptocurrencies with tables and graphs that fetched data every minute from REST API CoinGecko.
• CS 410 Data Engineering
- C-Tran Data Pipeline Project: Used GCP and Confluent Kafka to create a data pipeline for bus sensor events from a public transportation agency in Washington (C-Tran). Set it up to make daily requests, validate/clean data, store and integrate in PostgreSQL.
• CS 445 Machine Learning
- KiTS21 Challenge Project: Purpose of the project was to take the published kidney CT scans and train them on MLP, CNN and K-means models to predict segmentation scans showing background, kidneys, tumors, and cysts. I trained the CNN model on the scans reaching detection accuracy of 77%.