What is the ProjectVIEW MCP server? The ProjectVIEW MCP server is the connector that links AI assistants, Claude and ChatGPT, to ProjectVIEW ERP through the Model Context Protocol (MCP). It lets your people run headless operations without opening ERP screens, ask questions of their own data in plain language, create the dashboards and KPIs they need, and build their own AI agents and assistants. It is part of DANAOS ProjectVIEW AI and is available with ProjectVIEW ERP v12. ProjectVIEW ERP is the integrated ERP built on the BoQ ↔ WBS ↔ Cost Codes framework. The MCP server gives an AI assistant structured access to that data, so answers are tied to real budgets, commitments and approvals, not to a spreadsheet someone exported last week. What is MCP (Model Context Protocol)? MCP is an open standard for connecting AI assistants to business systems. A business system publishes what it can do (read data, run an operation, return a result) once, through an MCP server. Any assistant that supports MCP can then use those functions in a conversation. For a project-based company this matters for one reason: it removes the custom integration work between each AI tool and the ERP. The assistant does the language work. ProjectVIEW ERP stays the system of record that structures, validates and controls the data. What can you do with the ProjectVIEW MCP connector? The connector covers four capabilities: Run headless operations. Carry out ERP tasks from the assistant, without opening ERP screens. Ask questions of your own data in plain language. Query projects, costs, subcontractors, procurement and approvals the way you would ask a colleague. Create the dashboards and KPIs you need. Describe the view you want and build it from ERP data, instead of waiting for a report request. Build your own AI agents and assistants. Combine ERP data and instructions into assistants for recurring tasks in your department. Each capability works on the same unified data set. A question about committed cost, a KPI on margin and an agent that checks quotations all read the same BoQ, WBS and Cost Code structure. Which AI assistants does the ProjectVIEW MCP server support? The ProjectVIEW MCP connector is available for Claude and ChatGPT. Your teams can work in the assistant they already use and reach ProjectVIEW ERP data from there. Use the connector with your organisation’s business or enterprise workspace of Claude or ChatGPT, under your own agreement with that provider. That keeps the assistant inside your company’s account and data policies. What do example prompts look like? The examples below show the kind of request a user types and what the connector does in ProjectVIEW ERP. They are illustrative. What a user can see and do depends on configuration and on that user’s role. Prompt in Claude or ChatGPT What happens in ProjectVIEW ERP “Which projects are most likely to exceed budget?” Reads budget, committed cost and forecast data across the projects the user can access and returns the projects at risk, with the cost lines driving the gap. “What is the current committed cost for concrete on Project X?” Retrieves the committed cost recorded against the concrete items and Cost Codes of Project X. “Show me my pending approvals.” Lists the documents waiting for this user’s approval, as defined by the approval workflow. “Which subcontractors represent the highest commercial risk?” Reads subcontract data, certifications and variations and returns the subcontractors that need attention, with the reasons. “Build a KPI dashboard of committed vs actual cost by WBS for Project Alpha.” Creates the dashboard from Project Alpha’s WBS, commitments and actual costs. “Set up an assistant that flags quotations above the budget rate.” Builds an assistant that compares incoming quotations with budget rates and flags exceptions. Approval of any purchase stays with the responsible user. What are headless operations? Headless operations are ERP tasks carried out without opening ERP screens. The user states what needs to happen. The assistant passes it to ProjectVIEW ERP, which applies the same validation rules, workflows and authorisations as if the user had entered it in the screen. This is useful for people who do not live in the ERP: site managers, directors, approvers on the move. The principle is the one we apply to voice capture on site: “People speak. AI interprets. ProjectVIEW ERP structures, validates and controls.” Read more in how ProjectVIEW AI enables headless operations. How is access controlled? The connector works within ProjectVIEW user roles and permissions. A user reaches only the data and functions their role allows in ProjectVIEW ERP. Connecting an assistant does not widen that access. This follows the governance chain used across ProjectVIEW AI: Identity → Role → Permission → Data → AI → Action. AI should operate inside enterprise governance, not around it. Consequential actions always need human approval. See ProjectVIEW AI data security and governance for the full model. Why connect an AI assistant to the ERP instead of exporting data? Exported spreadsheets are out of date the moment they are saved, and they lose the links between budget, commitment and cost. ERP records what happened. AI explains why it happened and what will happen next. That only works when the AI reads the live, structured record. Answers trace back to BoQ items, WBS elements and Cost Codes. Every user works from the same data set, with the same permissions. No parallel copies of commercial data in personal files. For questions answered by the built-in assistant inside the ERP, see the ProjectVIEW AI assistant. For system-to-system integration, see the ProjectVIEW ERP API. When is the ProjectVIEW MCP connector available? The MCP connector was released with ProjectVIEW ERP v12 in October 2026. Version 12 is generally available to all clients. It also brings ProjectVIEW AI into daily processes and adds a new Production module for project-based manufacturing. Frequently asked questions Do I need ProjectVIEW ERP to use the MCP server? Yes. The ProjectVIEW MCP server connects AI assistants to ProjectVIEW ERP data. If your data sits in other systems, ProjectVIEW AI can also work on top of existing systems through APIs and integrations. Can the assistant change data in the ERP? Yes, through headless operations, within the user’s role and permissions and the ERP’s validation rules. Consequential actions always need human approval. Which assistants are supported? The ProjectVIEW MCP connector is available for Claude and ChatGPT. Is our data used to train AI models? ProjectVIEW AI retrieves ERP context at the time of the question, so no retraining of models on your data is required. How the connected assistant handles prompts is governed by your organisation’s plan and agreement with Claude or ChatGPT, so use a business or enterprise workspace. How is the MCP connector priced? Pricing depends on the modules, the number of users and the deployment option. Your DANAOS Projects contact will confirm the options for your setup. Want to see the MCP connector on your own data? Book a demo and we will show Claude or ChatGPT working on ProjectVIEW ERP data: questions, KPIs and headless operations. Book a Demo