🎓 EECS @ UC Berkeley
💻 Full-Stack Developer | Machine Learning Researcher
🌱 Exploring AI, computational biology, and scalable web applications

I am a Full-Stack Developer and Quantitative Analyst with a focus on building data-driven web applications. My expertise lies in bridging the gap between raw data and user experience—whether that’s engineering backend systems with Ruby on Rails or developing algorithmic trading strategies using Python and pandas.

I am passionate about writing clean, testable code (TDD with RSpec is a staple of my workflow) and translating complex mathematical concepts into functional software. Currently, I am focused on projects that leverage financial data to create actionable insights

Languages: Python, Java, SQL, JavaScript (React), C++, R, Bash
Frameworks & Tools: PyTorch, TensorFlow, Django, Flask, scikit-learn, OpenCV, Pandas, NumPy
Cloud & DevOps: AWS (EC2, S3, RDS, CloudFront), Docker, Supabase, GitHub Actions
Specialties: Machine Learning, Full-Stack Web Development, Data Engineering, Bioinformatics

• AI-driven playlist creation: Built a React Native mobile app that uses Ollama (LLaVA vision model) to analyze uploaded images and curate personalized Spotify playlists from a user's saved tracks.
• AI-powered music curation: Implemented a two-stage local AI pipeline using LLaVA for image context analysis and Mistral for semantic song selection, matching visual mood and aesthetics to specific music recommendations.
• Full-stack integration: Developed an Express.js backend with Spotify OAuth, image processing (Multer), and REST APIs to handle authentication, song selection, and playlist creation.
• Cross-platform mobile app: Built with Expo and React Native, featuring seamless Spotify authentication, image upload handling, and embedded playlist previews.
• End-to-end user experience: Designed a frictionless workflow: Spotify login → Image upload → AI analysis → Automatic playlist generation with custom cover art

Technologies: React Native, Expo, Express.js, Ollama (LLaVA, Mistral), Spotify Web API, TypeScript, Node.js

Full-Stack AWS-Deployed Game

• Browser-based roguelike with procedurally generated maps and AI-driven trainer battles
• Full-stack architecture: Django REST, React, PostgreSQL, deployed via AWS S3/CloudFront
• Integrated Supabase authentication and persistent session-based game state

Hackathon Project – Agentic AI for Gmail & Outlook

• Built a Chrome extension that generates replies, summarizes emails, and processes attachments
• Leveraged Google Gemini API for multimodal AI and tone-customized email generation
• Designed with a privacy-first approach and robust error handling for real-world use

Optical Music Recognition → MIDI

• Converts scanned sheet music into playable MIDI using CNNs, RNNs, and OpenCV
• Full-stack app with Flask + React, supporting uploads, segmentation, and real-time playback
• Designed scalable APIs and evaluation tools for symbol recognition and note accuracy

Machine Learning for Bioinformatics

• Applied Elastic Net Regression on RNA-Seq data to identify subtype-specific gene biomarkers
• Performed enrichment analysis using STRING and DAVID for biological validation
• Finalist @ ISEF 2022; selected participant in Regeneron STS 2023

• LinkedIn: linkedin.com/in/daniel-lin-mx777
• GitHub: github.com/Dragoknight777
• Email: dlin0505@berkeley.edu

⭐️ Thanks for visiting my profile! Check out my projects and feel free to connect or collaborate!