DrupalCamp Italy 2026 Speakers Preview AI Migrations, Civic Platforms and Enterprise Security

Ahead of DrupalCamp Italy's return to Bologna on 22 October 2026, six speakers have shared with The Drop Times how they are tackling complex challenges across real-world Drupal projects. Their written responses reveal the engineering behind more than 1,600 Italian public-administration portals, an AI-assisted migration that reduced 163 paragraph types to 42, and the infrastructure needed to deploy enterprise AI securely. The discussions also explore how organisations can modernise established Drupal systems without disrupting services and make everyday content editing easier.

These experiences reveal the difficult decisions behind technology projects that must operate at scale, accommodate years of legacy content and meet demanding security requirements. Artificial intelligence offers new possibilities, but it also introduces risks that developers must understand and manage. Across the speakers' responses, the focus remains on practical implementation, architectural trade-offs and lessons that other Drupal teams can apply.

DrupalCamp Italy 2026 will take place at the Savoia Hotel Regency in Bologna, bringing together Italian and international Drupal practitioners for technical sessions and community discussions. Organised by SparkFabrik, bmeme and Striped Codex, the event returns for its second edition following the revival of DrupalCamp Italy in Rome in 2025. The programme includes presentations in Italian and English on artificial intelligence, enterprise architecture, content management, migrations and open-source development.

One of the largest implementations featured at the conference is the Sportello Telematico Polifunzionale (STP), a public-service platform developed by Gruppo Maggioli. Alessandro Macchi, lead developer at Gruppo Maggioli, will explain how Drupal functions as the application core for more than 1,600 Italian public-administration portals. The platform processed approximately one million digital applications during 2025, according to the conference description and Alessandro's account, although the two sources differ on the precise total.

Alessandro's presentation, "The Public Administration Digital Ecosystem: Drupal as the Application Core," scheduled for 14:30–15:10 CEST in Room A, examines how Drupal coordinates complex civic services rather than simply managing content. One important principle is "once-only" data collection, which seeks to prevent citizens from repeatedly submitting information already available to public authorities. The approach requires integrating Drupal with national digital services while respecting the applicable legal and technical requirements.

The architecture uses Drupal's plugin system to build sequential processing pipelines for operations such as data validation, digital signatures and payment integrations. Synchronising information across 1,600 independent portals presents another challenge, particularly when simultaneous requests threaten to overload central infrastructure. Alessandro described combining asynchronous content ingestion through Google Cloud Storage with just-in-time loading of complex forms to reduce that pressure. Alongside these technical decisions, automatically populated fields and user-centred interfaces aim to simplify procedures for citizens and administrative staff.

Another substantial engineering challenge will be presented by Roberto Peruzzo, software developer at SparkFabrik. His session, "163 to 42: Dismantling and Reassembling a Drupal Migration with AI," examines a migration from Drupal 10 to Drupal 11 involving 163 paragraph types that needed to become 42 structured bundles. Some original paragraph types contained entire pages of mixed HTML inside individual fields, making conventional content mapping particularly difficult.

Roberto recalled feeling both concerned and curious when he first examined the source website. The problems extended beyond its complicated content model: the client did not realise that the database held information not visible on the live site, while the destination design continued changing during development. Manual investigation quickly became difficult to maintain. After a conversation with colleagues over dinner, Roberto began experimenting with AI not merely as a coding assistant, but as a tool for understanding the source system itself.

He developed an agent with small, focused capabilities that could explore the database, identify orphaned content and map source information to the evolving destination architecture. Analysis that previously required lengthy manual investigation could be repeated within minutes. The approach, previously presented at DrupalCon Rotterdam 2026, used 12 atomic agent skills alongside migration patterns for decomposing mixed HTML and reconstructing structured content. The team could then rebuild the destination pages through controlled migration processes rather than relying on AI to transform everything automatically.

Roberto emphasised that the results depended on carefully managing the agent's context and treating mistakes as risks to be caught by migration code. He considers AI a useful companion for understanding undocumented websites, but not a substitute for engineering judgement.

The client and development team were satisfied with the outcome, according to Roberto, although he continues to describe the method as experimental. He is now exploring similar techniques for a WordPress migration. His Italian-language session takes place from 10:50–11:30 CEST in Room A.

The security implications of increasingly autonomous AI systems will be addressed by Henk Beld, who works in strategic partnerships at amazee.io. His session, "Navigating the Edge: Why Data Sovereignty and Privacy Are Mandatory for Next-Gen Enterprise AI," is scheduled for 12:20–13:00 CEST in Room B. Henk argues that established security practices are struggling to keep pace with the speed of AI adoption.

He identified three risks associated with insufficiently controlled AI deployments: untraceable global data routing, execution without adequate sandboxing and uncontrolled expenditure from autonomous processes. His proposed infrastructure framework combines centrally managed service keys, private AI gateways configured for Zero Data Retention, Kubernetes container isolation and gateway-level spending limits. These safeguards can help organisations control where data travels, what automated processes can execute and how much they can spend. Henk's central argument is that security and compliance controls must be built into the infrastructure supporting AI development, rather than left to individual applications or added after deployment.

Architectural risk also lies at the centre of Adriano Cori's presentation. Adriano, a software engineer at bmeme, will examine how enterprises and public administrations can modernise established Drupal platforms without replacing everything at once. His session, "Enterprise Evolution Without a Big Bang: The Strangler Fig Pattern with Drupal," is scheduled for 17:05–17:45 CEST in Room A. According to Adriano, discussions about decoupling often overlook the practical difficulty of managing a transition while services remain operational.

The Strangler Fig pattern allows a modern frontend to run alongside an existing Drupal monolith, with a routing layer gradually transferring responsibility for individual routes. Both systems can operate under the same public domain while retaining the existing Drupal backend and data. Adriano described a method that supports incremental migration and relatively straightforward rollback if problems occur. However, route ownership, observability, caching, authentication, runtime resolution and failure handling all require careful planning. His session focuses on managing those responsibilities while protecting earlier software investments and avoiding a disruptive wholesale rewrite.

Questions about when and how to decouple Drupal will receive a different treatment from Vincenzo Gambino, Drupal architect at Striped Codex Limited. His session, "Decoupled Drupal in Practice," takes place from 10:50–11:30 CEST in Room B. Vincenzo believes teams often use the terms decoupled, headless and progressively decoupled interchangeably without fully understanding their architectural differences.

His demonstration will show how one Drupal content model can serve Astro, Next.js and a Python FastAPI service. The session covers Drupal core's JSON:API and serialisation capabilities, contributed modules such as Consumers and Simple OAuth, and tools supporting aliases, routing and custom endpoints. Vincenzo's main message is that effective content modelling matters more than simply choosing a new frontend framework. He also has a personal connection to the event: DrupalCamp Italy emerged from a conversation between him and Marco Marini of Sapienza University of Rome, followed by discussions with Stefano Mainardi of SparkFabrik and Claudia Astazi that helped turn the idea into a community gathering.

While several speakers focus on infrastructure and architecture, Nadiia Nykolaichuk, technical copywriter at ImageX, will address the everyday experience of Drupal content editors. Her session, "Practical Drupal Tweaks for Smarter, Faster, and More Enjoyable Content Editing," is scheduled for 17:05–17:45 CEST in Room B. Nadiia approaches Drupal from the perspective of a content professional who understands the technical decisions shaping editorial workflows.

Her presentation brings together practical improvements involving CKEditor toolbar customisation, scheduled publishing, automatic URL generation, content linking by title or keyword, real-time SEO feedback, accessibility checks and AI-powered assistance. The examples include established practices and lesser-known Drupal tools that can reduce repetitive tasks, prevent common mistakes and improve consistency. Nadiia hopes participants with different technical backgrounds will discover solutions they can implement in their own projects.

Beyond their technical presentations, the speakers also emphasised the value of meeting other Drupal practitioners. Alessandro  and Henk look forward to exchanging ideas about public services, open source and secure enterprise AI, while Adriano is particularly interested in comparing implementation challenges and architectural mistakes. Roberto welcomed the revival of the Italian Drupal community and expressed interest in Alex Moreno's keynote on the open web and Florent Torregrosa's session on Design Tokens API in Drupal core. Having previously spoken at DrupalCon Rotterdam, he is looking forward to presenting in Italian and discussing how AI is changing software engineering. Nadiia, whose speaking debut took place in Kortrijk, Belgium, is also looking forward to meeting contributors and exploring Bologna. Peruzzo additionally thanked the event's organisers and sponsors for their work.

DrupalCamp Italy 2026 will take place on Thursday, 22 October 2026, at the Savoia Hotel Regency, Via del Pilastro 2, Bologna, Italy. The programme brings together practical case studies, architectural discussions and community perspectives from Italy and abroad. Readers can find the complete schedule, speaker information and registration details on the official DrupalCamp Italy 2026 website.