← Back to Projects Financial Data Assistant with PostgreSQL + GPT
Let users query structured financial data in natural language and receive clear, application-ready responses.
Categories
GenAINLPML
Tech Used
PythonVector DBRAGLangChainPostgreSQLLLMFastAPIDockerAWSStreamlit
Problem
Business users often need insights from market data without writing SQL or manually translating questions into database operations.
Approach
- Stored market data in PostgreSQL and organized it for structured querying
- Combined retrieval and LLM reasoning to interpret user questions where useful
- Returned formatted responses designed for dashboards, APIs, and decision-support interfaces
Results
- Natural-language access layer for structured financial datasets
- Reusable pattern for database-backed AI assistants