🚀 Scalability: Implement Redis caching and API rate limiting

[shard-c6]

Description

As the user base grows and the consultant database expands past 500 species, we need to protect our backend and LLM APIs from abuse and reduce latency.

Tasks

• Introduce a Redis caching layer for the consultant API to cache exact-match queries (e.g., caching the response for 'Monstera watering schedule' for 24 hours).
• Implement API rate limiting (e.g., using Upstash or a local Redis instance) to prevent spamming the Gemini API.
• Add database connection pooling if not already configured to handle high concurrency.