About

Challenge

The management of funded projects is growing more diverse, more complex and more demanding, both nationally and internationally. Managing multi-project portfolios — especially under European programmes such as Horizon Europe and other multilateral funding instruments — is held back by highly fragmented processes, a lack of integration between tools, and a persistent dependence on outdated practices such as exchanging emails to communicate and to collect documentation. The picture is harder still in international consortia, which bring together entities under different regulations and administrative realities and therefore require high levels of coordination, transparency and compliance.

The members of the consortium have run into these structural gaps in their own day-to-day work. Consultancies see a model that is out of date, largely supported by generic or ill-suited software and by error-prone manual processes. Research support offices managing dozens of projects with very specific characteristics need an efficient digital solution to organise, archive and monitor every piece of data and documentation across a project’s lifetime — and find it difficult to consolidate technical and financial data in an integrated way, which undermines their ability to respond to the deadlines and targets set by funding bodies. At the same time, emerging technologies such as artificial intelligence remain under-used in this domain, leaving the potential of automation and prediction largely unexplored.

Solution

PMD-AI is a cloud platform organised in interconnected layers that bring together visualisation, processing, indexing and an AI engine.

The meta-dashboard is the central point for monitoring and interaction. It consolidates information from multiple projects into an integrated, real-time view, so that managers can track relevant variables — execution rates among them — identify the critical points that need immediate intervention, and reach detailed reports.

The processing layer hosts the specialised components: an AI-powered automatic configuration engine that adapts the platform to the specific needs of each project, plus modules for open data integration, automatic validations, document generation and predictions based on historical data. An open-source framework acts as the base infrastructure, handling communication between the processing, indexing and visualisation layers, keeping the solution modular and extensible, and letting the consortium concentrate its effort on the innovative components.

The approach is iterative and user-centred: every development step is grounded in real use cases and continuously validated with stakeholders, and user-centred design techniques are applied to keep the interface intuitive and effective.

Objectives

• O1. Develop an innovative digital platform that centralises and simplifies the management of funded projects, consolidating multi-project information in real time through an integrated meta-dashboard.
• O2. Implement intelligent automation of configurable templates, using artificial intelligence to pre-populate configuration models based on project typologies and reduce manual work.
• O3. Apply AI technologies to map relevant variables, predict critical events and generate actionable insights, enabling proactive and effective management of the resources allocated to projects.
• O4. Integrate technical and financial data from different sources and platforms, ensuring interoperability and the harmonisation of the information essential to project execution and monitoring.
• O5. Improve the decision-making capacity of the entities involved through advanced analytical and predictive tools that support real-time intervention and raise project execution and success rates.
• O6. Ensure the solution adapts and scales to diverse contexts, in both the public and the private sector, and can be deployed in national and international markets.

Innovation

Generic tools such as Microsoft Project, Asana and Trello already support scheduling, task assignment and real-time collaboration, but they are not designed for funded projects, where specific legal and documentary requirements apply. Closer competitors exist — EMDESK, Planisware and OpenProject — each with valuable features, and each with clear limits:

• A centralised meta-dashboard. EMDESK targets programmes such as Horizon Europe with planning, execution and reporting features. PMD-AI goes further with a global, real-time view over every project regardless of typology, allowing far broader and more effective monitoring.
• AI-driven configuration and validation. Planisware offers recognised strategic planning and execution tooling, but does not fully exploit artificial intelligence to map requirements and documents automatically, run automatic validations or suggest improvements from historical data. PMD-AI’s engine reads project documents — calls for proposals, technical documentation, Gantt charts — to identify critical parameters and generate tailored templates, flagging gaps and critical areas that need attention.
• Automatic template generation and a single documentation point. OpenProject is a flexible open-source solution supporting classic, agile and hybrid methodologies, but lacks AI-assisted generation of personalised templates and the centralisation of deliverables and documentation in one place.
• Predictions from historical data. Using AI, the platform provides predictions and actionable insights so that managers can work proactively, anticipating problems instead of reacting to them.

At a later stage, relevant players in the ecosystem — consultancies, for instance — will be able to build configuration templates whenever a new call is published and make them available to third parties, so that other entities can start from an already optimised template. Partnerships with the bodies responsible for managing calls may also allow “official” template versions.

Work plan

The work is organised into eight activities over 36 months: five of industrial research, one of experimental development, one of project management and one of dissemination.

Potential

The platform targets every kind of entity, public and private, that wants a digital tool to support the management of nationally or EU-funded projects, as well as the consultancies working in this area. SMEs are a particularly relevant segment: they would gain a competitive option for managing projects of this nature, where today they have few realistic alternatives.

The main areas of scientific and technological uncertainty concern the design of AI models able to process and interpret complex, heterogeneous data from sources such as calls for proposals, schedules and execution reports while keeping analyses accurate and reliable; the definition of effective methodologies to train models on anonymised and/or synthetic historical data that remain representative of the diversity of funded project typologies; and the integration of the platform’s components without compromising performance or the user experience.