Gemini API with LangChain Integration

LangChain is a popular framework for building LLM applications with composable chains, memory, and tool integrations. The langchain-google-genai package provides a first-class Gemini integration. ShopMax India uses LangChain with Gemini to power a customer FAQ bot that retrieves answers from product documentation using RAG.

The integration exposes Gemini as a ChatGoogleGenerativeAI object that plugs into any LangChain chain. You can combine it with ConversationBufferMemory for multi-turn conversations, vector stores for retrieval, and LangChain tools for function calling.

The below example shows how ShopMax India builds a product FAQ chain using Gemini and LangChain with retrieval augmented generation.

Running the RAG-powered FAQ chain,

It gives the following output,

Vector store built with 3 documents

Q: How long does it take to charge the Samsung Galaxy A55?
A: The Samsung Galaxy A55 takes 65 minutes to fully charge using a 45W fast charger.

Q: Can I return a product from Pune?
A: ShopMax India offers carry-in service at Mumbai, Bangalore, and Hyderabad. Pune customers
   should use the 7-day courier return option available through the website.

Q: Is EMI available on a Rs 8000 purchase?
A: Yes, EMI is available on orders above Rs 5000. Zero-cost EMI on HDFC and SBI cards.

For production at ShopMax India, replace Chroma with Vertex AI Vector Search for scalability, and persist embeddings in Cloud Storage. Use ConversationBufferWindowMemory with a window of 5 turns to keep the FAQ bot context-aware without unbounded token growth.