Alexandro Osornio
Software Engineer
Professional Summary
Software Engineer with 4+ years of experience developing and deploying deep learning solutions for production line automation. Skilled in Python, C#, Java, SQL, PyTorch, ONNX, and cloud integration via REST APIs. Proven ability to integrate AI models into production environments where previously manual quality control processes were transformed into fully automated systems, thus improving throughput while also increasing defect detection accuracy by more than 25%
Work Experience
ML Software Engineer
Advanced Spectral Technology | April 2023 - Present
- Automated quality control utilizing deep learning computer vision models with Tensorflow, PyTorch, and ONNX, for clients such as Intel and Western Digital, resulting in higher production line throughput with fewer defective products.
- Designed a custom CNN to detect 5-micrometer features on 8K images, improving inspection accuracy by over 20%.
- Integrated Python-trained neural networks into C# pipelines and deployed them within .NET Windows Forms applications for seamless operation.
- Automated robotic motion procedures for handling product parts and capturing inspection images using C# (.NET), eliminating manual intervention.
- Designed operator-friendly interfaces to facilitate continuous learning of deployed neural networks, reducing onboarding time for new designs and fine-tuning.
Software Engineer
ISAUSA | Jan 2022 - April 2023
- Automated classification of scanned document images by training and deploying deep learning neural networks using Keras and ONNX, achieving 96% accuracy in document categorization.
- Optimized inventory management by developing AI-powered image processing tools to locate, extract, and categorize labels on boxes from video streams, reducing manual sorting time by 33%.
- Developed a REST API and client-server architecture using Java and MySQL to enable remote access to resource-intensive AI technologies, enabling access on handheld devices and enhancing scalability.
Projects
Machine Learning for Sentiment Analysis
Software Engineer | August 2021 -- May 2022
- Designed and developed a sentiment analysis algorithm using Python (Keras), for classification of product reviews into multiple categories (e.g., positive, negative, neutral).
- Led a team to integrate the algorithm into a web application, enhancing its functionality and user experience.
- Implemented data preprocessing techniques (e.g., tokenization, stemming) and optimized the model to reduce computation performance.
Skills
Machine LearningPythonC#JavaJavaScriptSQLTypeScriptReactNext.jsGit/GitHubLinuxAWS.NETFastAPIPyTorchTensorFlow
Education
Bachelor's in Computer Science | GPA: 3.75
California Lutheran University | 2018 - 2022
Course Work:
- Machine Learning (Python)
- Data Structures and Algorithms (C++)
- Object-Oriented Programming (Java)
- Database Management (MySQL)
- 3D Graphics (Java)
- Robotics
- Software Design and Planning
Projects:
- Developed a machine learning model for image classification as part of the Machine Learning course, achieving 90% accuracy on the CIFAR-10 dataset.
Honors:
- Dean's List (2020--2022), Member of the Computer Science Honor Society
Skills
Machine LearningPythonC#JavaJavaScriptSQLTypeScriptReactNext.jsGit/GitHubLinuxAWS.NETFastAPIPyTorchTensorFlow