Hunjun Shin

10 projects · 2026

Projects

Hackathons, datathons, and applied ML work — from social science research tooling to real-time BCI systems.

Hackathon2026

Flanner

HackPrinceton Spring 2026 · Knot API track, 1st Place

A mirror on your delivery habits. Pulls six months of real DoorDash / Uber Eats orders via Knot API, decomposes each dish into ingredients with K2 Think V2, generates a weekly home-cooked plan respecting calendar and dietary constraints, and pushes the exact ingredients to a real Amazon Fresh cart in one tap.

  • Next.js 15 + React 19 on Vercel; FastAPI on Cloud Run; MongoDB Atlas
  • Knot API in production mode: TransactionLink for delivery history, AgenticShopping for real Amazon Fresh cart push
  • K2 Think V2 for weekly plan reasoning; Gemma 4 31B for photo recognition and check-in classification
  • Photon spectrum-ts bridge for live iMessage orchestration during the demo
Next.js 15React 19FastAPIPythonMongoDBK2 Think V2GemmaKnot APICloud RunVercel
Hackathon2026

MotZip

Next-Gen Hacks Beta · Spring 2026

Bilingual (EN/KO) restaurant discovery built around two ideas: voice-first filtering on a 3D map, and an AI phone agent that calls restaurants for you. Speak a query, the map sinks non-matching food into the ground; pick questions (reservations, vegetarian, wheelchair access...), click "Call N selected," and watch real Twilio calls return ✓/✗/? per question.

  • Next.js 16 + React 19 + Tailwind 4 frontend; MapLibre GL + Three.js for the 3D scene
  • FastAPI on Cloud Run; Google Places API for restaurant data, dedup'd across 7 cuisine groups
  • Voice search: Google Cloud STT → Gemini 2.0 Flash filter extraction → in-process filtering → Cloud TTS reply
  • Batch calls: Twilio chained <Gather> per question, per-turn STT + Gemini parsing, streamed back as a checklist
  • Graceful degradation: ElevenLabs Scribe/Turbo as STT/TTS fallback, keyword heuristics if Gemini JSON parse fails
Next.js 16React 19Tailwind 4MapLibre GLThree.jsFastAPITwilio VoiceGemini 2.0Google Places
Hackathon2026

BCI Neural Visualization System

Brain Storm BCI Hackathon · 1st Place ($1,000)

Real-time neural activity visualization and BCI array placement guidance system. Processes live data from a 1024-channel micro-ECoG array (32×32 grid) to help neurosurgeons optimize BCI array placement during surgery. Hosted by Precision Neuroscience at Microsoft NERD Center.

  • CNN-based denoising (ResNet U-Net with CBAM attention)
  • Kalman filter and EMA-based position tracking
  • Real-time heatmap smoothing with configurable web UI
  • 500 Hz sampling, ~20–50 ms end-to-end latency
PyTorchResNet U-NetCBAMKalman FilterReal-timeECoG
Datathon2026

Zerve × ODSC AI Datathon

ODSC AI Datathon · 2026

End-to-end MLOps pipeline as a 32-block, 42-edge parallel-converge DAG inside the Zerve canvas: schema/leakage validation → EDA + 15-stage funnel → a 5-candidate AutoML pool → drift detection → champion picked for serving → weekly retraining loop. A Next.js frontend renders the DAG and calls a deployed FastAPI for live inference, reading canvas variables in real time.

  • 5-candidate model pool: calibrated XGBoost + RF + HGB soft-vote ensemble, PyTorch tab-MLP, sklearn GBM — isotonic calibration throughout
  • PR-AUC 0.2645, ROC-AUC 0.812 on 3.5M rows / 17,541 users (1.84% upgrade rate, ~53:1 imbalance); top-5% precision ~9× lift
  • Tree SHAP + segment-level diagnostics map top features to marketing actions
  • Weekly drift watch (PSI 0.10/0.25 + KS) with append-only event store, label-stable retraining gates
  • Next.js 14 + ReactFlow frontend draws the canvas DAG and queries the deployed FastAPI for live PNGs and inference
PythonXGBoostPyTorchscikit-learnSHAPFastAPINext.js 14ReactFlowZerve
Hackathon2026

Booky

YHack 2026 · with Hyoungseo Son

Social reading platform that transforms solitary reading into a connected experience. Readers highlight passages, discuss with friends, make story-branching choices (Detroit: Become Human style), and explore an AI-generated solar system visualizing reading compatibility.

  • Next.js 16, React 19, Three.js for 3D planet visualization
  • FastAPI backend with Firebase Firestore and ChromaDB
  • Google K2 Think V2 for AI content, Gemini 2.0 Flash for voice Q&A, Vertex AI Imagen for illustrations
Next.js 16React 19Three.jsFastAPIFirestoreChromaDBVertex AI
Project2025

Reflecta

LLM Goal-Setting Chatbot · Top 12 LikeLion 2025 Hackathon

An LLM-based goal-setting and conversational AI system for daily reflection. Achieved Top 12 in LikeLion 2025 Hackathon. CHI 2026 Poster submission planned.

  • FastAPI backend, deployed on GCP Cloud Run with Docker
  • Conversational goal-setting and tracking flow tuned for daily check-ins
  • Vercel-hosted demo frontend
PythonFastAPIGCPDockerLLM
Project2024

PantryPilot

End-to-End MLOps Recipe Assistant · 2nd Place, Google Cambridge MLOps Expo 2025

Full-stack MLOps pipeline for personalized recipe generation: LLM fine-tuning (Llama 3.2 3B with LoRA), DPO-based feedback loop for continuous model improvement, data pipeline with Great Expectations validation, Airflow orchestration, and observability with Slack alerts. Deployed finetuned LLM on GCP Cloud Run for scalable inference, with GCS + DVC for data and model versioning. Features dietary restriction enforcement, admin dashboard, and automated model registry.

  • Llama 3.2 3B fine-tuned with LoRA, then DPO loop for continuous model improvement
  • Airflow-orchestrated data pipeline with Great Expectations validation and Slack alerts
  • GCP Cloud Run deployment + GCS / DVC for data and model versioning
  • Admin dashboard, automated model registry, dietary restriction enforcement
FastAPIReactPyTorchLoRADPOAirflowGCPDockerGreat Expectations
Project2025

MNIST MLOps Pipeline

Automated Retraining with User Feedback

Complete MLOps pipeline for MNIST digit recognition with automated model retraining based on user feedback. Features interactive drawing canvas, real-time predictions, and continuous model improvement workflow.

  • FastAPI inference service deployed via Vertex AI on GCP Cloud Run
  • GitHub Actions for CI/CD across train, evaluate, and deploy stages
  • React frontend with interactive drawing canvas and feedback capture
FastAPITensorFlowGCPVertex AIGitHub ActionsReactDocker
Project2024

FullStackGPT

LangChain + Streamlit AI Apps

AI-powered web applications using LangChain — DocumentGPT (document Q&A with RAG over FAISS), QuizGPT (automatic quiz generation), and PrivateGPT (local LLM for private document analysis).

  • RAG pipeline with FAISS vector search and OpenAI GPT
  • Streamlit frontends with chat history and document upload
  • Local LLM variant for fully offline private document analysis
LangChainStreamlitOpenAI GPTFAISSRAG
Project2024

Job Board

Flask + MySQL Full-Stack Platform

Full-stack job board platform with job search, company profiles, and real-time recommendations. Features MySQL database integration with SQLAlchemy ORM and EDA / ML analysis.

  • Flask backend with SQLAlchemy ORM over MySQL
  • Job search, company profiles, and recommendation surfaces
  • EDA and ML analysis on collected job listings
FlaskMySQLSQLAlchemyPython