Add Dozer LLM banking chatbot sample

[Iineman2]

/claim getdozer/dozer#1690
/claim #1690

Summary

• Add a complete usecases/llm-banking-chatbot sample based on the Dozer + LLM banking personalization article.
• Configure Dozer over multiple local CSV datasets: customers, accounts, transactions, and card products.
• Expose raw APIs plus a joined customer_profile API, then consume that context from a LangChain chatbot.
• Add Chroma-backed product retrieval, deterministic local embeddings, an optional OpenAI chat path, and an offline fixture mode for quick reviewer verification.
• Add runnable docs, Docker Compose, sample data, env template, and root README discoverability.

Demo

Short demo video: https://github.com/Iineman2/dozer-samples/releases/download/dozer-1690-demo/dozer-1690-demo.mp4

Verification

• python -m py_compile usecases\llm-banking-chatbot\app\banking_chatbot.py
• CSV header/row validation for all four fixture datasets
• .\usecases\llm-banking-chatbot\app\.venv\Scripts\python usecases\llm-banking-chatbot\app\banking_chatbot.py --offline-fixture --customer C001 --question "Which card is best for travel and dining rewards?"

The offline verification returns Ava Sharma's customer context, ranks Dozer Travel Signature first for travel-heavy spend, and generates a personalized recommendation. Docker is not installed on my local Windows host, so the Dozer container path is documented but not locally executed here.

Related bounty issue: getdozer/dozer#1690