Production MCP Servers

Your team needs an answer, but the information is spread across different systems. Instead of checking each one, they could ask ChatGPT or Claude: “Which customers still have open orders?” I build the connection that lets the assistant look up the answer in your data, with access permissions in place. That’s where MCP servers come in.

Try it with real data.

Just try it for yourself: explore my projects or look up market data. This browser demo uses prepared answers based on portfolio content and live server data.

MCP in action

Choose a question below to see the data behind the answer. To ask your own questions, connect ChatGPT or Claude just below the demo.
Ask Daniel
Markets
Connect to your ChatGPT or Claude

Use this public, read-only demo in your own ChatGPT or Claude account. Add the server URL below, enable the connection in a conversation and ask your own questions.

ChatGPT

In ChatGPT on the web, enable Developer mode under Settings → Security and login. Open Plugins, choose + and add this URL with No Authentication. Then enable the connection in your chat.

https://mcp.danielsolves.ai/mcp

Requires developer-mode access; your workspace may need to enable it.

Official setup guide ↗

Claude

Open Customize → Connectors → + → Add custom connector and paste this URL. Then enable it from the + menu in your chat.

https://mcp.danielsolves.ai/mcp
Official setup guide ↗

From a question to an answer.

Your team should be able to ask a question without hunting through dashboards or asking someone to run a report. I connect your AI assistant to the information it needs.

1. Ask a question

Your team asks their AI assistant a question about the business.

2. Look up the right information

The connection I build looks up the answer in your systems, with access permissions checked first.

3. Get an understandable answer

The assistant explains what it found using the data returned by your systems.

Your data. Your access rules.

You decide who can see what

People can only access the information they are allowed to use. The public demo above uses public data only.

You can follow what was accessed

A record of each request shows who asked for which information and when.

Tech stack.

TypeScriptNestJSMCPOAuth 2.0Docker

Want AI to work with your data?

Tell me which systems you use and what your team would like to find out. I’ll help you choose a useful first connection and work out who should have access. A rough idea is enough.