Thank you!
We will contact you shortly
AI in construction is moving beyond general-purpose tools for generating text and processing documents. It is increasingly becoming a practical way to work with real project data and support construction project management.
We continue to expand the AI Assistant in Dedal CPMS. Previously, it helped users work with estimates and pricing. Now it can also analyze construction project performance and answer questions about the current status of projects.
There is no need to navigate through multiple reports to find the required figures. Users can simply ask questions in natural language, such as “How is the project doing?”, “Where are we over budget?” or “Who owes whom?”
The AI Assistant accesses the Dedal CPMS data available to the user and generates its response based on that information.
One of the simplest use cases is to ask the Assistant for an overview of the current financial status of a particular construction project.
For example:“How is the Sunny Residential Complex doing?”
The response can include the project's key indicators:
● plan;● budget;● forecast cost at completion;● actual costs;● variances;● margin.
This allows users to quickly understand the overall project situation without having to review several reports and metrics separately.It is a practical example of AI in construction project management: artificial intelligence works with actual data from a specific project in Dedal CPMS rather than with generic information.
For companies managing several construction projects at the same time, the Assistant can analyze the entire project portfolio available to the user.
For example:
“Which projects need attention?”
The Assistant provides an overview of the projects the user is authorized to access and displays their key financial indicators. Projects with the most significant issues appear at the top, while margin performance is presented using clear traffic-light indicators.
This helps users identify which projects require attention and then move directly to a more detailed analysis of a particular project.
For companies managing multiple sites and projects, this approach simplifies construction data analytics and provides a clearer view of performance across the entire project portfolio.
A high-level variance is not always enough. To take action, project teams need to understand what is happening within the project.
Users can therefore ask:
“Where are we over budget on this project?”
The system identifies the work items with the largest variance between forecast and plan.
This makes it easier to move from the overall financial result to the specific activities affecting it and supports more effective construction cost control.
The analysis can then go one level deeper.
If a particular work item shows a variance, the user can immediately ask the Assistant to explain the underlying reason.
For example:
“What caused the cost overrun?”
The AI Assistant breaks the work item down to the resource level and compares planned and actual quantities and prices.
This makes it possible to see what contributed to the variance — for example, whether more material was consumed than planned or whether the actual purchase price was higher than expected.
As a result, users receive not only an indication that a variance exists, but also the underlying data needed to understand what caused it.
This is an important aspect of construction automation: the system helps users move quickly from a high-level indicator to the individual factors behind it.
Another use case is analyzing settlements with contractors and clients.
Users can ask:
“Who owes whom?”
The Assistant can display information on contractors and clients, including:
● certified work / acts;● payments;● outstanding balances.
The information can be requested for a selected period or for the entire project history.
This provides a quick way to review settlements without manually collecting information from multiple documents.
The AI Assistant can also answer questions about inventory and warehouse balances.
For example:
“What is left in stock?”
The response can show remaining resources, including quantities, total value and average price.
A single conversational interface can therefore be used to work with different types of Dedal CPMS data — from project financials to resources, inventory and settlements.
One of the main advantages of this approach is that users do not have to know exactly where a particular metric or report is located in the system.
They can ask the question in the same way they would ask a colleague:
“How is the project performing?”
“Where is the largest cost overrun?”
“What caused it?”
“How much do we owe the contractor?”
This does not replace the reports, documents or project management tools available in Dedal CPMS. Instead, the AI Assistant provides another way to access and analyze existing project data — through conversation.
It demonstrates how AI for construction can be integrated not as a standalone general-purpose chatbot, but as part of specialized construction management software.
Along with the new capabilities, we have updated the Assistant's system rules.
When generating an answer, it must use the retrieved figures correctly and specify the relevant currency and data date. If the required information is unavailable, the Assistant states that clearly instead of substituting a different figure.
This is particularly important when working with financial data, where an incorrectly interpreted metric can change the meaning of the entire analysis.
The AI Assistant follows the access permissions configured in Dedal CPMS.
It can only access projects and data that the individual user is authorized to view.
For example, if an employee only has access to selected projects, the Assistant cannot be used to retrieve information about other company projects.
The AI Assistant also operates in read-only mode: it can analyze and present information but cannot modify data in Dedal CPMS.
The development of artificial intelligence in construction is not limited to generating text, drawings or images. One of its most practical applications is helping teams work more efficiently with the large volumes of project data they already collect.
This is the direction in which we are developing the Dedal CPMS AI Assistant.
It can now help users quickly assess project performance, identify work items with cost overruns, investigate the reasons behind variances, review contractor and client settlements, and check inventory balances — all based on data already available in the system.
This is another step toward digital transformation in construction, where information about plans, actual performance, resources and finances is not simply stored in a system but becomes easier to access, analyze and use for day-to-day management decisions.
The AI Assistant is a built-in chat assistant in Dedal CPMS. It answers questions about the company's construction project data, including estimates, resources and prices, planned and actual project performance, project portfolios, settlements and inventory balances. It can also provide guidance on how to perform specific actions in the software.
Administrator setup
Go to:
Settings → Maintenance → Services → AI Service
The Assistant operates through an external AI service that is connected separately for each company. The configuration is available only to the company administrator.
1. Obtain an API key from your preferred provider:
● Anthropic (Claude) — create a key in the developer console at console.anthropic.com under API Keys. The key begins with “sk-ant-”. Please note that a Claude.ai Pro/Max subscription cannot be used for API access. A separate API key is required, and the account must have sufficient funds for API usage.● Google (Gemini) — create an API key in Google AI Studio.● OpenAI (GPT) — create an API key on the OpenAI platform API keys page. A ChatGPT subscription does not provide an API balance; API access and billing are configured separately on the OpenAI platform.
2. On the Services page, select the provider. The default model will be selected automatically, but it can be changed if required.
3. Paste the key into the API Key field and click Save Key.
4. Click Test. The system will send a test request to the selected provider and display the result.
The API key is stored in encrypted form. Once saved, the key itself is never displayed again; only the “Key configured” status is shown.
To replace an existing key, simply save a new one. To disable the service, save an empty field.
Once configured, the AI Assistant becomes available to company users, while access to project data continues to be controlled by their Dedal CPMS permissions. The cost of AI requests is charged according to the API pricing and billing terms of the selected provider.
Available OpenAI models
Dedal CPMS supports the following OpenAI models:
The availability of a specific model depends on the OpenAI account. You can verify access by clicking Test in the AI Service settings.
AI usage control
If required, the administrator can set a monthly token limit under Settings → Services.
Current usage is displayed in the chat. Once the configured limit has been reached, the AI Assistant stops processing new requests.
The monthly limit is disabled by default.
Would you like to get free advice on business automation?