Tanzu Platform Chat: AI Chat Client for Cloud Foundry

What is Tanzu Platform Chat: AI Chat Client for Cloud Foundry?

Tanzu Platform Chat (cf-mcp-client) is a Spring chatbot application built with Spring AI, designed to be deployed to Cloud Foundry and consume platform AI services. It works with Large Language Models (LLMs), Vector Databases, and Model Context Protocol (MCP) agents.

How to use Tanzu Platform Chat: AI Chat Client for Cloud Foundry?

Deploy the pre‑built Spring application to Cloud Foundry using mvn clean package and cf push. Bind available AI services (e.g., GenAI‑based LLMs, embedding models, Postgres vector databases, or an MCP server) with cf bind-service and restart the application with cf restart ai-tool-chat. For vector database support, upload a PDF via the document tool; the chatbot will then answer queries about the file.

Key features of Tanzu Platform Chat: AI Chat Client for Cloud Foundry

• Deployed and run on Cloud Foundry with cf push.
• Integrates with GenAI‑compatible LLM service instances.
• Supports binding to vector databases (Postgres with embedding models).
• Uses Model Context Protocol (MCP) agents via user‑provided services.
• Persists chat memory across restarts when bound to a vector DB and embedding model.
• Accepts PDF document uploads for contextual question‑answering.

Use cases of Tanzu Platform Chat: AI Chat Client for Cloud Foundry

• Provide a conversational AI interface that uses Cloud Foundry’s GenAI tile for LLM responses.
• Enable document‑based Q&A by uploading PDFs and querying them via a vector database.
• Extend AI capabilities by connecting to external tools through MCP agents.
• Maintain conversation history across application restarts and scaling events.

FAQ from Tanzu Platform Chat: AI Chat Client for Cloud Foundry

What platforms or cloud environments does it require?

The application is designed exclusively for Cloud Foundry environments that offer the GenAI tile or other LLM services. It also works with standard Postgres databases and user‑provided MCP servers.

How do I bind an LLM to the chatbot?

Create a GenAI service instance (e.g., cf create-service genai [plan-name] chat-llm), bind it with cf bind-service ai-tool-chat chat-llm, then restart the application.

Can I add vector database support?

Yes. Bind an embedding model service and a Postgres instance, then restart. After binding, you can upload a PDF via the document tool to enable document‑based queries.

How do I connect an MCP agent?

Create a user‑provided service with the endpoint URL (cf cups mcp-server -p '{"mcpServiceURL":"https://your-mcp-server.example.com"}'), bind it, and restart. The LLM will then be able to invoke the agent’s capabilities.

Does the chatbot remember previous conversations?

If both a vector database and an embedding model are bound, chat memory persists across application restarts and scaling.