Jan 2024 - Jun 2024 | Seattle, WA
ML Research
As part of my university capstone project with TE Connectivity, I spearheaded machine learning application into their expander line process.
Self-teaching myself machine learning through research, hands-on practice, and rigorous study, I applied my skills to real-world problems.
Leveraging tools such as TensorFlow and PyTorch, I implemented complex algorithms including Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), specifically Long Short-Term Memory (LSTM) networks, to optimize the expander line process.
These models resulted in a ~20% reduction in scrap material, significantly decreasing the dependency on operator intuition and saving hundreds of engineering hours.
Throughout the capstone project with TE Connectivity, I took ownership of implementing key machine learning solutions that directly impacted the expander line process.
Working closely with their engineers, I ensured our collaboration was seamless, combining my technical expertise with their operational knowledge to drive meaningful results.
I proactively guided my team, facilitating problem-solving discussions and setting a clear direction for integrating machine learning into existing workflows.
Model Performance Analytics
To further enhance operational efficiency, I developed a custom analytics dashboard using Python, Streamlit, and MySQL.
The webpage provided real-time insights into machine parameters, integrating predictive machine learning models that anticipated key performance indicators.
This reduced manual intervention, streamlining workflows by ~30%.
Throughout the project, I utilized my full stack knowledge—ranging from JavaScript for frontend components to Streamlit for deployment—to ensure a seamless integration of these AI-driven solutions.
My ability to self-learn and independently apply machine learning to solve industry-specific problems significantly contributed to improving product quality and reducing operational costs.
By promoting a strong focus on quality and performance, I helped streamline operations, reduce material waste, and improve overall productivity.
This collaborative effort not only showcased my technical contributions but also strengthened the relationship between our team and TE Connectivity, ensuring the successful delivery of innovative solutions.
Highlights:
University of Washington
2022-2024 | Seattle, WA
BS in Electrical & Computer Engineering
- Algorithms & Data Structures
- Database Systems
- Fundamentals of ML Optimization
- ML for Sig Processing Applications
- Probability and Stats for Engineers
- Digital Circuits and Systems
- Advanced Multivariable Calculus
Teaching staff for an ECE class at UW
Coming soon: Jorgo - Kyrgyz Programming Language
A programming language designed specifically for Kyrgyz speakers that is a simple translation of English syntax.
Started as a joke with my friends, I finally had the courage to start working on it.
Jorgo aims to make coding feel natural and intuitive in the Kyrgyz language, allowing native speakers to express programming concepts in their own tongue.
- Python-like implementation with Kyrgyz syntax and semantics
- Designed to capture Kyrgyz linguistic patterns and thought processes
- Includes beginner-friendly documentation in Kyrgyz
- Open-source project to promote tech literacy in Kyrgyzstan
Allstack: AI Models Comparison
It's like Reddit for AI models—users can compare models, upvote their favorites, and add comments.
There's a new AI model coming out almost every two weeks, which makes it hard to keep track of the best options.
This platform solves that by letting you easily compare models side-by-side.
I plan to add more lists for AI agents, copilots, and include even more models.
It's not fully ready yet—I know I should launch soon but life and chasing perfection gets in the way.
Check it out!
Disobedient Robots
This project, led by Nicolás Kisic Aguirre, explores the idea of robots as creative partners instead of just tools.
It uses advanced tech to make robots more independent, interactive, and capable of working with humans in new and artistic ways.
Checkout my essay for more.
- Autonomous Navigation: Robots can independently move through spaces without manual control.
- Obstacle Avoidance: Equipped with LiDAR and depth cameras, they detect and avoid obstacles in real-time.
- Environmental Mapping: Robots create real-time maps of their surroundings for better navigation.
- Flexible Adaptation: Can adjust to different scenarios, spaces, and experimental conditions.
- Advanced Sensing: Utilize sensors to perceive and respond to dynamic environments effectively.
- Collaborative Behavior: Designed to act as creative collaborators, not just tools, in artistic and experimental projects.
Theia Tech
Theia Technologies is an award-winning American optics design and manufacturing company known for quality and innovation. This was made during my internship and I worked on the Lens Selector and Calculator pages. Below you can see a screenshot of my work.
- Over 650 active users
- Enables customization when choosing lenses
- Supports in selecting lenses using a lens calculator and a resolution simulator
- Overall a highly efficient and user-friendly filter
Live Demo
Carpool App
Developed and launched an iOS app from the ground up, managing all phases of the project as a freelance developer. The app is now discontinued.
- Lightning-fast scrolling on the home feed screen with an impressive 60 FPS performance
- Improved trip search efficiency
- Seamless and smooth user interface
Internships
Stealth Startup
Summer 2023
[ Learn More ]
Realm AI was a startup that would revolutionize the home buying process by utilizing open-source Large Language Models (LLMs) and creating a single place resource for new home buyers. I conducted in-depth research and integrated advanced LLMs to elevate the startup's technological framework.
R&D Engineering Intern
-
Fine-tuned the model that achieved a 90% success rate and successfully passed the real estate exam which qualified to get the real estate agent license in over 40 states.
-
Implemented advanced LLMs, driving a ~20% increase in predictive analytics accuracy and a ~15% performance boost, significantly enhancing user satisfaction and achieving key business goals.
Webkey Spring, Summer 2022
[ Learn More ]
Webkey is a full-service web design firm specializing in dynamic, database-driven websites, collaborative project management tools, and online publishing. I worked on multiple projects where I did full-stack web development.
Software Engineering Intern
-
Developed and maintained a complex web application using Python/Django for the backend and JavaScript/React for the frontend, resulting in a ~30% increase in user engagement and a 20% reduction in load times.
-
Implemented RESTful APIs and optimized database queries, resulting in a 25% improvement in system performance and supporting over 10,000 concurrent users.
-
Implemented multiple custom-made plugins using the Django framework to automate 100+ features of a large CMS application, which boosted efficiency and reduced operational costs.
-
Stack: Python/Django, JavaScript, PostgreSQL, HTML, CSS, React/Typescript, Linux,
Carpool
Summer 2019
[ Learn More ]
At Carpool, I built and launched an iOS app from scratch, handling all aspects of the project as a freelance developer.
Software Engineering Intern
-
Developed and launched a scalable iOS carpool app, improving trip search efficiency by 40%, which reduced user search time and increased user engagement by 30%, leading to higher app retention and satisfaction.
-
Improved scrolling performance of the home feed screen reaching 60 FPS which enabled the customers to enjoy a smooth user experience.
-
Stack: Swift, SwiftUI, RxSwift, Xcode, Firebase, Realm
