Hello, my name is
Tim Choy.
I love building scalable software and understanding how systems work under the hood.
From distributed systems to AI-driven applications, I am passionate about exploring new technologies and turning what I learn into elegant, reliable solutions. I'm especially interested in backend engineering and designing the infrastructure behind the next generation of data-driven applications.
About Me
Hi, my name is Tim! I am a recent UC Berkeley graduate with degrees in Computer Science and Data Science. Most of my experience with programming has come from building real applications, breaking them, and figuring out how to make them better. Engineering for me is about stretching my understanding of how things work under the hood, especially when there's room to optimize, scale, or rethink the architecture entirely.
Throughout my engineering journey, my work has spanned across systems programming, data engineering, and AI research. Whether it's designing distributed architectures, building data pipelines for bioinformatics, or training emotionally-aware conversational LLMs, I strive to build elegant solutions while learning something new along the way. What motivates me most is tackling complex engineering problems that require both strong fundamentals and a bit of creatvity.
Skills & Technologies
Languages & Frameworks
Tools & Technologies
Featured Projects
Here are some of my recent projects that showcase my skills and experience
AI-Powered Second Brain
I built a full-stack, AI application that enables private, locally hosted LLM interactions. I designed a RAG pipeline that ingests personal documents and provides context-aware responses while ensuring complete data privacy. For the backend, I implemented a high-performance FastAPI service supporting document processing, vectorization, and semantic search using ChromaDB, along with SQLite-backed chat history persistence. On the frontend, I developed a responsive React + Electron interface featuring real-time message streaming, seamless file upload for document ingestion, and intelligent autocomplete for document references using custom hooks and shared context. Together, these components form a fast, privacy-preserving desktop assistant capable of leveraging a user’s knowledge base for intelligent local querying.
Smart Calendar App
I developed a full-stack, cross-platform smart calendar application that allows users to schedule and manage events in real time across multiple devices. To ensure flexibility and maintain data consistency, I designed a type-safe calendar system in TypeScript that supports a variety of entry types and can be easily extended with new features. I implemented secure authentication, role-based access control, and granular permission settings to enable collaborative scheduling, calendar sharing, and efficient task management. Built with React Native/Expo and Supabase for the backend, the application delivers a seamless, reliable, and secure experience for both individual and team productivity.
Distributed Filestore
I designed a distributed file storage system deployed on AWS and using Go. The system splits files into chunks, stores them across multiple EC2-based storage nodes, and tracks metadata in DynamoDB. A public API server handles upload/download requests, proxies chunk transfers to private nodes inside a VPC, and manages node health via heartbeat monitoring. Infrastructure is fully provisioned using Terraform, including EC2 instances, networking, IAM roles, and DynamoDB tables. The result is a scalable, cloud-native storage architecture that demonstrates distributed systems design, backend engineering, and AWS automation..
PintOS
Over the course of a semester, I worked within a large, low-level C codebase to implement core components of an operating system. I contributed to the design and development of user program support, multithreading, and a minimal yet fully functional file system. Key problems I tackled included implementing process-control system calls, building a user-level synchronization library, designing a priority-based CPU scheduler, and developing an inode-based file system with persistent storage. The project required extensive debugging, testing, and collaboration with other students to ensure correctness and robustness at the kernel level. Together, these contributions produced a functional, multithreaded operating system capable of running user processes and managing a reliable file system.
Work Experience
Deep Learning Researcher
Project AEI
- ▸Designed, optimized, and trained conversational AI models for emotional intelligence that enhanced memory efficiency and training speed using PyTorch and Vertex AI
- ▸Synthesized research on state-of-the-art deep learning architectures to guide model design, involving closed-form continuous-time layers, transformer encoder layers, and residual connections
Data Engineering Intern
Triplebar
- ▸Designed and implemented an ETL pipeline to extract, transform, and load data from Postgres database into a graph-based model, enhancing data representation and analysis for genetic relationships
- ▸Built an interactive dashboard for bioinformatics teams using Dash, supporting efficient graph traversal and visual exploration of protein strain data
- ▸Implemented statistical simulations with probability distributions to generate synthetic genetic data, improving the accuracy and robustness of genotyping pipelines
Mobile Developer Intern
Socale
- ▸Maintained cross-platform mobile app codebase with rigorous unit testing and documentation in Flutter and Dart
- ▸Collaborated with designers to develop dynamic, customizable user profiles with interactive components
- ▸Implemented Firebase services for user authentication, real-time chat, and cloud-based data storage
Let's Connect!
I'm always open to conversations, collaboration, and new opportunities — whether it's about engineering, AI, or something completely unexpected. Feel free to reach out if you'd like to work together, ask a question, or just say hello.