In the field of public construction tendering, preparing a competitive tender is a complex process. One of the most labor-intensive components is creating the project schedule – the Gantt chart. This is, essentially, a visual map of a project – a timeline spanning the entire project, covering every activity, from pouring the foundation to finishing touches, as well as details on the sequence of work, estimated durations, allocated resources, and dependencies between activities.

Using traditional planning methods, creating such a chart requires days of work: collecting data from estimates, establishing the logical sequence of tasks, calculating durations, allocating teams and equipment, and checking for resource overlaps – all with significant risks of error. The eConfaire Immersive Design team is working on a solution that will fundamentally change this process.

In the context of public tenders in Romania, as stipulated by Government Decision No. 1/2018, the project schedule is a mandatory document that must be aligned with the technical and financial proposals. The figures in the cost estimate, the activities in the schedule, and the amounts in the payment schedule must be fully consistent with one another. Any inconsistency may result in the disqualification of the bid.

Automating a process that takes days to complete

The solution developed by eConfaire will be integrated into Enterprise Construction Flows, a platform used by Leviatan Group specialists for the efficient management of internal processes. The basic principle is simple: the solution will automatically retrieve data from the estimates already created in the platform and generate a complete Gantt chart, without any manual intervention.

Specifically, the extraction of hundreds or thousands of items from a quote, their organization into a logical structure, the estimation of durations, and the establishment of dependencies will be performed automatically. A static quote will be transformed into a dynamic execution plan, featuring temporal relationships between activities, resource allocations, and realistic duration estimates.

How it works: specialized AI agents for each stage

The solution is based on a suite of AI agents. An AI agent can be understood as a “virtual specialist” trained to solve a specific type of problem. Each stage of the process will have its own agent: one will validate the input data, another will calculate the duration of activities taking into account the scope of work and applicable standards, and a third will establish logical relationships between activities – it will understand, for example, that you cannot install electrical systems before erecting the walls, or that certain finishing works can proceed in parallel.

This modular approach will allow for much greater accuracy than a single general-purpose algorithm could provide, and the result will be a schedule that reflects the reality on the construction site.

Flexible adjustment and intelligent resource allocation

The solution will offer several options for adjusting the total project duration, including an advanced non-linear adjustment method. Instead of proportionally reducing the duration of all activities, the system will optimize each activity differently, considering the nature of the work and actual technical constraints.

Additionally, the solution will include an intelligent resource allocation mechanism that automatically detects overlaps – situations in which the same resources are scheduled in two different locations at the same time.

Estimated results and impact

By using the solution developed by eConfaire, the time required to generate a Gantt chart will be reduced by approximately 70%. Thus, a process that currently takes days will be almost entirely automated.

The solution will not eliminate the human factor but will reposition it. Instead of spending time on repetitive work, Leviatan Group specialists will focus on validating and refining the results – a concept known as “Human in the Loop.” The expert’s role will remain essential, but it will be oriented toward quality decisions, not mechanical operations. At the same time, the ability to quickly generate multiple planning scenarios will provide a real competitive advantage in tenders.