Portfolio project for DraftKings Analyst II, RG Analytics role (Job ID: jr13280).
# 1. Review project context
cat CLAUDE.md
# 2. Start Claude Code
claude
# 3. First command
claude "Review CLAUDE.md and list first 3 development tasks"Generate a demo DuckDB with a majority of Critical/High/Medium cases:
python scripts/seed_demo_db.py --players 400claude/ # Claude Code configuration
├── skills/ # Domain knowledge (5 files)
├── agents/ # Specialized personas (5 files)
├── workflows/ # Multi-step procedures (3 files)
└── context/ # Reference data (4 files)
dbt_project/ # Data transformation
ai_services/ # LLM integration
backend/ # FastAPI REST API
frontend/ # React dashboard
tests/ # Testing
docs/ # Documentation
- Snowflake (target) + DuckDB (dev) + dbt + Python + FastAPI + React + OpenAI API
See docs/PORTFOLIO_WRITEUP.md for a concise, recruiter-ready overview of the project.
- Persisted analyst queue with lifecycle tracking (Not Started → In Progress → Submitted)
- Full Case File workbench with HITL notes, SQL evidence, and AI assist transparency
- Read-only SQL execution with logging + Snowflake-safe guardrails
- Regulatory trigger checks (MA/NJ/PA) logged as deterministic SQL evidence
- Manager-grade Analytics dashboard (throughput, rigor, compliance signals)
- DraftKings-authentic tech stack + responsible AI framing
- CLAUDE.md - Project overview
- claude/skills/ - Domain expertise
- claude/agents/ - Specialized tasks
- claude/workflows/ - Step-by-step procedures
- docs/HITL_REVIEW.md - Completed analyst review with SQL queries
- docs/WEEK9_REPORT.md - Live integration + analytics + SQL workflow updates
- docs/PORTFOLIO_WRITEUP.md - Recruiter-ready project summary
- docs/TECHNICAL_DEEP_DIVE.md - Full technical rationale + AI workflow design
Add your final assets here when ready:
- Live demo link
- Walkthrough video (2–3 min)
- Sample PDF report