How does ProjectVIEW AI work? DANAOS ProjectVIEW AI works by retrieving live ERP data, structured by BoQ ↔ WBS ↔ Cost Codes, at the moment a question is asked or an agent runs, and passing it to the model best suited to the task. ProjectVIEW AI is the AI add-on to ProjectVIEW ERP: agents and an assistant that work on ERP data, combining domain-specific small language models with leading LLMs. Every consequential action goes to a person for approval. In short: deterministic ERP plus probabilistic AI. The ERP keeps the record. The AI reads it, explains it and suggests what to do next. What data layer does ProjectVIEW AI use? ProjectVIEW AI uses the data layer of ProjectVIEW ERP, the integrated ERP built on the BoQ ↔ WBS ↔ Cost Codes framework. Every cost, commitment, quantity and progress record is linked to a BoQ item, a WBS package and a cost code. That link is what lets AI compare like with like across estimates, purchase orders, subcontracts and actuals. ProjectVIEW AI can also read data from outside the ERP: Third-party ERP such as SAP and Oracle. Schedules from Primavera P6 and MS Project, and BIM models (.ifc). EDMS, documents, email, contracts and IoT data. Unstructured data is mapped to BoQ, WBS and cost codes, so a clause in a contract or a fuel reading from a machine lands on the same structure as the budget. See native connectors and the BoQ software behind the structure. Why retrieve data at inference instead of training on it? ProjectVIEW AI retrieves the relevant ERP context at inference time, when the question is asked or the agent runs. Models are not retrained on client data. This has three practical effects: Answers use today’s figures, not the figures at the date of the last training run. Role permissions are applied at retrieval, so the model only receives what the user is allowed to see. Client data stays client data. It does not become part of a shared model. What happens when you ask ProjectVIEW AI a question? The user asks in plain language, by text or voice, in the ERP or through the MCP connector. ProjectVIEW AI identifies the user and the role, and checks permissions. It retrieves the permitted ERP records and documents that relate to the question, linked by BoQ, WBS and cost codes. The model chosen for the task analyses the records and drafts the answer. The answer is returned with the figures and records it is based on, so the user can check it and drill into the ERP. Which models does ProjectVIEW AI use for which task? ProjectVIEW AI is model-flexible. The model is chosen per use case. Task Model type Example Pattern recognition and numerical analysis Domain-specific small language models (SLMs) Detect a quotation above the budget rate, abnormal fuel use, or a cost trend that points to overrun Reasoning, language and multi-step planning Leading LLMs Explain why a project is over budget, draft a briefing, plan the steps of an investigation Retrieval Semantic search on ERP records and documents Find the contract clause, the comparable BoQ item or the related variation Details are on the ProjectVIEW AI models page. How does a ProjectVIEW AI agent work, step by step? Every agent follows the same cycle: Detect. The agent watches its process and spots a deviation, for example a quotation above the budget rate. Investigate. It pulls the related records: the budget line, past rates, the supplier history, the WBS package. Explain. It states in plain language what is off and why it matters. Recommend. It proposes an action, such as negotiating, re-tendering or reallocating budget. Request approval. It sends the recommendation to the person whose role owns the decision. Execute. After approval, the action is carried out in the ERP. Audit. Every step, input and decision is logged. The eight purpose-specific agents are listed on the ProjectVIEW AI Agents page. How is every AI action governed? AI should operate inside enterprise governance, not around it. Each request passes through one chain: Identity → Role → Permission → Data → AI → Action. Identity: who is asking. Role: what the person is responsible for. Permission: which projects, records and approvals the role allows. Data: only the permitted records are retrieved. AI: the model works on that data only. Action: anything consequential is executed only after human approval. Data is not sent to public consumer AI services. On Microsoft Azure, prompts are not used for model training, and data residency options, encryption at rest and customer-managed keys are available. See security and governance. Where does human approval sit? Human approval sits between Recommend and Execute. ProjectVIEW AI can detect, investigate, explain and recommend without waiting. It cannot commit cost, change a contract value, certify progress or post a financial entry on its own. Those decisions stay with the people who hold the authority in the ERP approval workflow. How does the voice pipeline work? Voice is designed to let site teams work headless, without opening ERP screens. Availability for your deployment is confirmed during scoping. The pipeline is designed to run in six steps: Voice capture on site. Transcription to text. Semantic interpretation of what was said. Association to BoQ items, WBS packages and cost codes. ERP fields populated, subject to validation and authorisation. Reconciliation with existing records. It is designed to produce a speech-to-text daily site log for the user to review. Voice recordings alone do not establish contractual entitlement or certified progress. People speak. AI interprets. ProjectVIEW ERP structures, validates and controls. Frequently asked questions Does ProjectVIEW AI need our data to train its models? No. ProjectVIEW AI retrieves ERP context at inference time, so no retraining of models on client data is required. What happens if our ERP data is incomplete? The answer is only as good as the data. AI will not fix broken processes. It will amplify them. ProjectVIEW AI shows which records it used, so gaps are visible and can be corrected. Can ProjectVIEW AI work with SAP or Primavera P6? Yes. ProjectVIEW AI can read data from SAP, Oracle, Primavera P6, MS Project and BIM, mapped to BoQ, WBS and cost codes. Can an agent execute an action without approval? Not a consequential one. Agents request approval from the responsible person before execution, and every step is logged for audit. Want to see it on your own data? See the agent cycle and the governance chain run in a ProjectVIEW ERP demo environment. Book a Demo