Tessitura MCP Servers
Custom MCP servers that connect your Tessitura to Claude, Claude Code, ChatGPT, Gemini, and your internal tools, with or without authentication, read and/or write access, and granular control over every action.
Your team already uses AI. Now it can use Tessitura.
Staff across your organization are asking Claude, ChatGPT, and Gemini for help every day, but those assistants can't see your sales, your patrons, or your season. So people copy and paste exports, or go without.
We build a custom MCP server for your Tessitura: a secure bridge that gives the AI tools your team already uses live, controlled access to the data and actions you choose. Ask a question in plain English and get an answer from your real numbers, in seconds.
- AuthSSO / OAuth · API key · Public
- AccessReadWrite
- ScopePer tool · per role · per field
See it in action
Staff working in their AI assistant, connected to their organization's Tessitura MCP server. Every tool call is permission-checked, every change needs approval, and everything is logged.
Animated demo: a development director asks an AI assistant to find lapsed donors and add follow-up notes, approving each write; a marketer checks Nutcracker sales and builds an email list.
Illustrative examples. Constituent data is fictional.
Granular control, your way
Every server is designed around your organization's policies. You decide who can connect, what they can see, and what, if anything, they can change.
With or without sign-in
Open and anonymous for public season information, API keys for your systems, or OAuth sign-in through your existing identity provider (Microsoft Entra ID, Google Workspace, Okta) for staff.
Read, write, or both
Start read-only. Add write actions such as notes, contact updates, or seat holds only where they make sense, with human approval required before changes are made.
Per tool, role & field
Expose only the tools each team needs, and redact sensitive fields like payment details, birthdays, or private notes so they never reach an AI model.
Audit logs & limits
Every tool call is logged with who asked, what ran, and what changed. Rate limits and query caps protect your Tessitura instance from runaway requests.
Example permission model
| Tool | Public | Box office | Marketing | Development | Admin |
|---|---|---|---|---|---|
| get_season_calendar | Read | Read | Read | Read | Read |
| get_performance_availability | Read | Read | Read | Read | Read |
| get_sales_summary | Not exposed | Read | Read | Read | Read |
| search_constituents | Not exposed | Read | Read, redacted | Read | Read |
| get_contribution_history | Not exposed | Not exposed | Not exposed | Read | Read |
| create_constituent_note | Not exposed | Write | Not exposed | Write | Write |
| update_constituent_contact | Not exposed | Write | Not exposed | Not exposed | Write |
| reserve_seats | Not exposed | Write | Not exposed | Not exposed | Write |
One server, three audiences
Use one server for everything, or separate servers per audience. Either way, each connection gets exactly the access it should have.
The public
Read-only tools over public information: season calendar, performance details, availability, and pricing. AI assistants can answer questions about your shows accurately, straight from the source.
Your staff
Each person signs in with their work account and gets the permissions of their role, inside Claude, ChatGPT, Gemini, or Claude Code.
Your systems
Internal apps, automations, AI agents, and our Tessitura AI Chatbot use the same governed tools, so there's one integration to maintain.
What your team can ask
No reports to build, no exports to wrangle. A few examples of what becomes a one-line question:
“What's still open in the orchestra for Saturday's matinee?”
“Find this patron's upcoming orders and resend their confirmation.”
“Which productions are pacing behind last season at two weeks out?”
“How many first-time buyers from the fall haven't returned yet?”
“Brief me on this donor before tomorrow's lunch.”
“List lapsed $1,000+ donors and when we last spoke to each.”
“Summarize this week's ticket revenue against budget by production.”
“Compare subscription renewals with the same point last year.”
“In Claude Code: check how price types are set up for Nutcracker, then update our TNEW template to match.”
Connected in a minute
Once your server is live, adding it to an AI tool takes one step. Staff add it as a custom connector in Claude, ChatGPT, or Gemini and sign in with their work account. Developers add it to Claude Code with a single command.
We provide setup guides for each tool and train your team on getting the most out of it.
# Add your Tessitura MCP server to Claude Code
$ claude mcp add --transport http \
tessitura https://mcp.your-theater.org/mcp
✓ Added HTTP MCP server tessituraHow we work
- 01
Discovery
We map the questions and tasks your teams want AI help with.
- 02
Tool & permission design
We define each tool, who can use it, and what stays private.
- 03
Build & test
We develop against your Tessitura test environment first.
- 04
Deploy & connect
We go live in your cloud or ours, and connect your AI tools.
- 05
Train & evolve
We train your staff and add new tools as needs grow.
Questions
What exactly is an MCP server?
MCP (Model Context Protocol) is an open standard, introduced by Anthropic in 2024 and now supported by OpenAI, Google, Microsoft, and others, that lets AI assistants use external tools and data safely. An MCP server is the piece in the middle: it defines exactly which actions an AI can take (like “check availability” or “add a note”), and handles the connection to Tessitura so the AI never touches your API directly.
Which AI tools can connect to it?
Any MCP-compatible client, including Claude, Claude Code, ChatGPT (via custom connectors), Gemini-based tools, and developer tools like Cursor and VS Code. Because it's an open standard, the same server keeps working as your team's tools change. It can also power your own internal apps, agents, and our Tessitura AI Chatbot.
Can we keep it read-only?
Yes, and we recommend starting there. Write tools (like adding notes or updating contact details) are opt-in, can be limited to specific roles, and can require explicit human approval before anything changes in Tessitura.
Is our patron data sent to AI companies?
Only what a tool returns, and only to the AI provider your team chooses to use, under that provider's terms. Business and enterprise plans from the major providers generally exclude your data from model training. Because we control the server, we can also redact sensitive fields such as payment details, birthdays, or private notes so they never leave it.
What do we need on the Tessitura side?
Access to the Tessitura REST API (the same API your website and TNEW integrations use) and a dedicated service user with appropriate permissions. We build and test against your Tessitura test environment before anything touches live data.
Where does the server run?
Wherever you prefer: in your own cloud (Azure, AWS, or Google Cloud), or hosted and maintained by us. Either way you get monitoring, audit logs, and updates as Tessitura and the AI tools evolve.
The Tessitura AI Chatbot
Give patrons the same live Tessitura answers on your website, 24/7.
Ready to get started?
Let's talk.
Whether it's AI on your Tessitura data, a website redesign, an SEO strategy, or custom software, we'd love to hear about it. Consultations are free, and we typically reply within 24 hours.
- Phone
- +1 (202) 800-0207
- Office
- 2021 L St. NW, Suite 101-341
Washington, DC 20036 - What to expect
- Free consultation · Custom solutions · Ongoing support