feat: preserve and display all CSV columns including transaction references

fix: remove false positive security checks for legitimate column names

feat: auto-detect merchant column from common aliases

upgraded chat UI with markdown rendering, currency highlights, and improved AI card styling

feat: enable versatile AI responses for creative and analytical queries

feat: improve UI/UX with modern chat interface and better visual hierarchy

fix: retrieval to pass all transactions for aggregation queries and filter display table by question context

fix: ensure accurate metadata handling and prevent missing-key errors

fix: ensure metadata keys for amount, date, merchant on all docs and avoid missing metadata error

Refactor QA chain and RetrieverService for improved clarity and functionality; update docstrings and enhance error handling

Integrate RetrieverService and enhance QA chain; initialize session state for embedding manager and retriever service

Refactor embedding creation process to use batch processing and persist path; implement RetrieverService for document retrieval

Enhance user interaction with question input UI and improve embedding management; update docstrings for clarity

updated success message for processing completion and cleaned up docstrings in ingestion module

Remove txt.py from repo