UC Davis Data Science student building backend systems, data tools, LLM-agent workflows, and iPad-native note-taking products.
I am currently looking for AI Engineer, Software Engineer, Data Science, Data Engineering, and Data Analyst internship opportunities.
Agentic data reliability platform where an OpenAI-compatible LLM chooses profiling, quality-check, memory, and report-building tools over uploaded business CSV or read-only PostgreSQL data before producing a structured, evidence-backed answer.
- FastAPI backend
- pandas-based profiling and checks
- LLM tool calling and prompt-designed JSON output
- Business CSV upload with owner, primary-key, and schema context
- Read-only PostgreSQL demo path for realistic business tables
- Redaction, retries, timeout handling, cost estimation, and deterministic guardrails
- LLM output evaluation against deterministic findings and evidence references
- Docker, GHCR container publishing, GitHub Pages demo, and GitHub Actions CI
Public evidence:
- Demo: https://sunnnn2005.github.io/data-quality-agent/
- Recruiter evidence pack: https://github.com/sunnnn2005/data-quality-agent/blob/main/docs/application-evidence-pack.md
- AI Engineer review card: https://github.com/sunnnn2005/data-quality-agent/blob/main/docs/ai-engineer-reviewer-card.md
- Open AI reviewer request: sunnnn2005/data-quality-agent#26 (comment)
- Public evidence health: 96/96 checks passing
- Test suite: 218 passing tests
- Current honest traction baseline: 0 confirmed external users, 0 accepted external feedback items, 0 GitHub stars
Repo: https://github.com/sunnnn2005/data-quality-agent
Native iPad note-taking app that turns handwritten headings into a navigable table of contents for student lecture notes.
- SwiftUI
- PencilKit
- Vision OCR
- Multi-page handwritten notes
- Folder-based organization
Repo: https://github.com/sunnnn2005/smart-handwritten-note-taking-app
- Languages: Python, SQL, R, Swift
- Data: pandas, NumPy, matplotlib, seaborn, regression analysis
- AI: LLM tool calling, OpenAI-compatible APIs, prompt design, structured JSON output, LLM evaluation
- Backend: FastAPI, Pydantic, REST APIs
- Tools: Docker, GitHub Actions, Git, GitHub
- iPadOS: SwiftUI, PencilKit, Vision OCR
- LinkedIn: https://www.linkedin.com/in/shun-lu-20060b210
- Email: shunlu2005@gmail.com