01.2 Energy & Utilities

In the energy sector, the ongoing transition is not limited to the shift to renewable sources alone, but is increasingly influenced by the decentralisation of energy systems and the volatility of markets. Reply operates across the entire value chain, supporting operators in generation, trading, networks and sales in the integrated management of physical assets and commercial relationships. The approach combines domain expertise with technological accelerators based on IoT and edge computing, as well as with pathways for adopting generative AI applied to business processes. This integration allows the complexity of the energy system to be converted into a measurable competitive advantage.

In the crucial transition from centralised production to a distributed model based on renewable sources and storage systems, Reply leverages operational data as a strategic asset, developing solutions that integrate AI agents to normalise information flows and optimise trading activities on spot, intraday, balancing, and ancillary services markets. Through advanced forecasting and specialised risk models, the solutions developed by Reply ensure regulatory compliance and support decision-making in contexts of high price volatility. This analytical capability improves the balance between supply and demand, strengthening the financial and operational resilience of energy companies.

In critical infrastructure, Reply ensures grid reliability through digitalisation, which involves both transmission and distribution system operators. The use of robotics, autonomous drones, and computer vision for system inspections reduces risks to personnel and ensures continuous monitoring, while edge computing enables the automation of operational processes. Sensorisation of high- and medium-voltage grids allows for real-time adaptation of infrastructure. Predictive analytics allows for the early detection of anomalies and overloads, improving the stability of smart grids and extending the lifespan of systems.

The solutions developed for the retail market transform the end customer from a simple consumer to an active participant in the energy system: by supporting the creation of engagement tools and the personalisation of commercial offers, Reply accelerates innovation in value-added services along the entire supply chain. Reply’s expertise extends to the management of decentralised energy resources through Virtual Power Plant solutions and demand response platforms that coordinate charging infrastructure and electric vehicles. Reply supports the integration of mobility into the electricity system with vehicle-to-grid technologies and enables dynamic pricing models for sector operators.

01.3 Financial Institutions

Reply is an active part of the entire financial services ecosystem (banks, insurance companies, wealth and asset managers, payment operators, consumer credit companies, mobility finance, regulators) and supports the integration of AI as a lever for transforming business models, customer experience, and industry productivity. The approach is not limited to automating existing processes: Reply enables the transition from procedural and sequential models to collaborative paradigms in which human and artificial agents cooperate in real time to solve complex problems. Conversational interfaces are progressively replacing traditional workflows, allowing specialists to access corporate knowledge, data, and legacy algorithms through agentic AI, which makes knowledge immediately usable and processes radically more agile.

Reply supports financial institutions in adopting AI in a pragmatic and effective way both through the application of technology for the intelligent automation of established processes and in rethinking service models. Financial institutions are no longer simply building proprietary portals and apps but are adapting to be present where their customers are, integrating with conversational agents who increasingly mediate access to digital services. In the insurance industry, for example, this is radically changing the customer experience: from requesting a quote to reporting a claim, interaction becomes more natural, contextual, and personalised through conversational interfaces.

In 2025, Reply deployed multi-agent architectures to orchestrate complex decision-making and operational processes by coordinating networks of specialists (humans and AI agents) working in parallel on highly complex tasks, such as risk profile analysis, wealth management advisory support, or credit assessment. Intelligent automation also extends to ICT factories, where AI agents support the entire software development cycle and infrastructure management, significantly increasing productivity and reducing delivery times. Reply also supports financial institutions in their technological modernisation programmes, helping them overcome the limitations of legacy systems and transition to flexible, scalable, and secure digital infrastructures, enabling the integration of advanced AI capabilities in full compliance with industry regulatory frameworks.

Reply supports financial institutions in implementing responsible AI frameworks, ensuring transparency, interpretability, and regulatory compliance in automated decisions. AI-based solutions also automate complex reporting and data consolidation processes for liquidity and capital risk management, significantly reducing the analysis time required by supervisory authorities. Significant attention is paid to the integration of ESG criteria, supporting financial institutions in sustainability reporting and alignment with new regulatory standards, also using AI for the analysis and continuous monitoring of sustainability metrics.

01.4 Government & Healthcare

Reply collaborates with government institutions and local public administrations in several countries to support the transformation of citizen services into efficient and interoperable digital ecosystems. Through the modernisation of information systems and the development of dedicated vertical solutions, Reply helps optimise and facilitate cooperation between agencies and user access to public and healthcare information. Its ability to manage highly complex, large-scale initiatives is reflected, for example, in the support provided to several national drug safety authorities through the implementation of cloud architectures designed to ensure the resilience of critical processes.

Over the course of 2025, Reply expanded its expertise in process digitalisation in areas such as the UK judiciary and key government ministries. It also expanded its offering to support the protection of critical infrastructure and technological sovereignty, working alongside the armed forces, public institutions, and healthcare organisations. Particular attention was paid to strengthening cyber resilience through the adoption of AI-based technologies that enable continuous monitoring of strategic assets and strengthen threat prevention and critical infrastructure security management processes.

In the healthcare sector, Reply supports organisations in leveraging AI to simplify the management of care protocols and regional and national regulations, which are characterised by high complexity and frequent updates. Machine learning solutions, for example, make it possible to automatically analyse regulatory documentation and promptly identify the actions required to ensure compliance. This approach contributes to more effective service planning and the maintenance of high standards of accountability, security and data protection, promoting conscious and sustainable digital innovation.

At the same time, AI enables the creation of advanced tools to enhance clinical information, such as intuitive dashboards that summarise data from medical records while keeping the physician at the centre of the decision-making process and contributing to more effective and safer healthcare. In this context, the advanced processing of large volumes of biomedical and radiomic data represents a central element of the clinical innovation enabled by Reply: through specialised platforms such as X-RAIS Reply, healthcare professionals are supported in the interpretation of diagnostic images, such as X-rays and CT scans, improving the accuracy of early diagnosis. Moreover, solutions such as Ticuro Reply enable continuous patient monitoring and the personalisation of therapeutic pathways.

01.5 Logistics

In the logistics sector, AI is evolving the way processes and operations are designed, managed, and executed across the entire supply chain. The growing complexity of physical and information flows, the integration of physical and digital channels, and the spread of next-generation automation systems now require platforms capable of consistently coordinating people, technologies, and operational decisions. In this context, Reply supports logistics operators and companies operating in various industries in evolving their supply chain models toward more efficient, resilient, and adaptive structures.

In warehouses and distribution centres, Reply’s algorithms support picking optimisation, dynamic inventory management, and resource allocation, while computer vision solutions monitor operational flows and identify anomalies in sorting and handling processes. Artificial intelligence is also used for demand analysis, capacity planning, and transportation management, contributing to more accurate and timely decisions even under complex operational constraints.

These applications help overcome fragmented operations management, enabling coordinated governance of logistics processes based on data, insights, and predictive models. Orchestrating flows along the supply chain improves operational precision, resource utilisation, and the ability to adapt to unforeseen events, with positive impacts on business performance and the overall sustainability of logistics and production activities.

The technological cornerstone of Reply’s logistics offering is the LEA Reply modular platform, designed to manage complex logistics processes in international contexts. Recognised by several market analysts as a global benchmark for logistics and warehouse management systems, LEA Reply currently supports leading companies in the automotive, fashion, retail, and food & beverage sectors, guiding them in the transformation of their supply chains towards more integrated and sustainable operating models.

01.6 Manufacturing

Reply positions itself as a strategic partner for manufacturing companies with a strong focus on innovation, enabling the development of advanced cognitive manufacturing solutions based on agentic AI. Through intelligent platforms, Reply enables the evolution of the entire production lifecycle, fostering autonomous and context-aware decision-making across different industrial domains. By combining technological excellence with deep process expertise, Reply develops agentic platforms designed to rapidly enable intelligent, scalable and reusable use cases. The goal is to transform production units into interconnected and data-driven environments, capable of overcoming information fragmentation and optimising production capacity, quality, and maintenance, ensuring full visibility into overall plant performance.

For granular control of factory operations, Reply offers modular solutions that integrate consolidated experience with leading market application platforms for ERP, MRP, PLM, Planning and Forecast with its proprietary Brick Reply solution, an advanced MES/MOM system for managing the execution, planning, and orchestration of production processes. In 2025, Brick Reply industrialised native agentic AI capabilities to accelerate decision making processes and enable a deeper understanding of production phenomena through predictive analysis, contextual suggestions and adaptive automation. In parallel, Axulus Reply scales Industrial IoT initiatives, governing frameworks, workflows and reusability in complex production environments. Thanks to AI, Brick Reply and Axulus Reply can operate synergistically, enabling them to query complex data in natural language, obtain immediate insights, and trigger operational and strategic actions directly in the field.

Managing physical flows to support production lines is another cornerstone of Reply’s offering. Through proprietary and market-leading solutions, enhanced by intelligent and autonomous features, Reply enables efficient production logistics, ensuring timely supply and end-to-end component traceability. Dynamic optimisation of supply and storage reduces inefficiencies and waste, maintains alignment with production plans even in variable scenarios, and improves operational and energy sustainability.

Production improvement is supported by the adoption of edge AI and computer vision technologies, which bring intelligence directly to the production lines, enabling real-time analysis and decision-making. This approach enables high-precision automated quality controls, capable of identifying defects that cannot be detected manually, and predictive maintenance solutions that intercept early signs of failure, reducing plant downtime and improving operational continuity and overall efficiency.

01.7 Retail & Luxury

In the innovation journey shaping the retail, fashion and luxury sector, Reply supports leading global brands throughout the entire transformation process, from redefining the customer experience to evolving operating and business models. By combining deep vertical knowledge of the retail, fashion and luxury industries with end-to-end technological expertise and a long-term architectural vision, Reply enables brands to operate in an increasingly omnichannel, data-driven and experience-centric environment, without compromising identity, exclusivity or service quality.

Reply supports retailers in the transition from fragmented models to integrated omnichannel ecosystems, where e-commerce, physical stores and mobile operate seamlessly, customer data is shared and decisions become faster and more contextual. This lays the foundation for the transition from simple automation to “smart” retail, where technology becomes a competitive differentiator.

In 2025, Reply introduced a new paradigm based on agentic AI, overcoming the limitations of traditional solutions in customer service and contact centres, clienteling and sales force support, inventory, dynamic pricing, logistics and replenishment, operations, and back-office. AI is a key element in enhancing both the end customer experience and internal efficiency, introducing conversational e-commerce and natural language product search, dynamic interaction personalisation, automatic creation of content consistent with the brand tone, and real-time support for sales and customer service.

Evolution of the AI Market in the Retail Industry

Source: Reply Research “AI for Retailers” - Reply analysis of PAC data relating to 12 countries (Italy, Germany, United Kingdom, USA, France, China, India, Brazil, Belgium, Netherlands, Poland, Romania); data in millions of euros

A further differentiating element of the Reply offer is the integration of AI with 3D technologies, digital twins and immersive environments, enabling advanced configuration scenarios and real-time customisations, supporting virtual try-ons, digital showrooms and immersive pop-ups and extending the concept of endless aisle in a natural and engaging way. The combination of generative AI and 3D allows brands to move from a simple product visualisation to a narrative and interactive experience, where customers can explore, customise, and understand the item’s value before purchasing.

01.8 Telco & Media

Reply supports leading European telecommunications companies in diversifying their portfolios and transitioning to new sustainable business models, supporting them in the design of processes and application architectures based on artificial intelligence. In 2025, Reply further strengthened its presence in infrastructure areas, with specialised expertise in network engineering, operations, testing, and validation to support network evolution.

The multi-agent architectures designed by Reply enable efficient development of Business Support Systems (BSS) and Operations Support Systems (OSS), creating flexible solutions and optimising integration with legacy systems. The adoption of AI fosters the creation of new services and the enhancement of existing ones: agentic AI architectures integrate automated tasks and workflows, allowing operators to quickly respond to customer needs. Customer care services are becoming increasingly conversational, thanks to the use of specialised AI models, which enable seamless and personalised interactions while improving operational efficiency.

Reply supports telcos’ transition to a software-based operating model by transforming the network access layer into distributed, cloud-native, and disaggregated edge platforms. The complex geopolitical landscape makes the adoption of a “security by design” approach essential: Reply ensures the resilience of critical infrastructures by integrating cybersecurity directly into the network architecture, protecting strategic assets and enabling operators to evolve into technology hubs capable of managing the entire stack securely and reliably.

In the media sector, Reply supports publishers in modernising their operating models through the development of agentic AI frameworks aimed at increasing productivity and time-to-market. Reply experts support publishing companies in creating innovative content, particularly for social media and advertising, optimising the time and costs of editorial processes such as copywriting, audio generation, and video production. This integration accelerates production and allows publishers to offer hyper-personalised content based on user preferences and real-time data analysis.

02.AI-POWERED SERVICES

Having become a structural component of new enterprise architectures, artificial intelligence lies at the core of Reply’s services, enabling the development of new processes, products and business models capable of generating tangible value for clients.

02.1 Knowledge Management

AI is profoundly changing the way companies organise, analyse, and leverage their information, both structured and unstructured. Knowledge management is no longer limited to simple document storage or search but is evolving toward the construction of dynamic knowledge ecosystems capable of supporting complex reasoning and operational decisions. Reply has invested in the specialisation of AI models to build vertical knowledge networks, specific to each client and each industrial context: the primary objective is to enable AI agents to understand the business domain, interpret technical and business documents, integrate multimodal input, and transform fragmented content into coherent logical entities.

The use of increasingly verticalised large language models for industry enables the development of advanced conversational systems capable of extracting, reaggregating, and redistributing knowledge in a contextualised form. Through integration with vector databases and Retrieval-Augmented Generation (RAG) architectures, corporate data is transformed into structured semantic representations, offering a conceptual vision of different knowledge domains. Based on this foundation, Reply further strengthens the effectiveness of its models through targeted fine-tuning activities, specialising them on the client’s application domain and priority use cases. This allows it to go beyond simple contextual information retrieval, improve consistency and reasoning, and ensure more precise alignment with the company’s actual processes. This approach is complemented by the adoption of continuous learning logics: each interaction and process generates new structured knowledge, fuelling the progressive improvement of models and execution policies. The result is a dynamic system capable of adapting to the company’s actual operational needs over time, while optimising performance quality and cost sustainability.

The evolution of knowledge management is thus resulting in the digitisation of entire workflows: the adoption of specialised agents is leading to the creation of new application architectures, in which hybrid teams, composed of human operators and agents, collaborate on complex processes. In the financial sector, for example, specialised agents can automate the management of commercial documents, support regulatory compliance, and synthesise complex reports into operational insights. In customer service, agents are progressively specialising thanks to the historical database of interactions, improving the quality of responses and reducing resolution times.

The ecosystem of partnerships and proprietary platforms allows Reply to integrate open-source, multimodal, and on-premises models based on the specific needs of client companies, ensuring high standards of security, regulatory compliance, and operational sustainability. The development and progressive industrialisation of agentic solutions were supported throughout 2025 by strategic partnerships with leading hyperscalers and new agreements with international technology leaders, including OpenAI. Reply also consolidated its proprietary Neurons Reply platform, evolving it into a central infrastructure for the creation, management, and orchestration of agentic AI solutions within the enterprise. Neurons Reply enables users with varying skill levels to aggregate, customise, and orchestrate agents to support business processes within a secure, compliance-ready framework.

02.2 Prebuilt AI Apps

The traditional model of enterprise software use, based on the configuration of large, monolithic applications like ERP and CRM that impose linear processes and static interfaces, is evolving towards agentic AI architectures. In this new setup, legacy systems retain their role as data providers, while interaction management is handled through agents: the “application” is no longer a closed product that predefines every operational step, but becomes a dynamic composition of services in which the interface and execution logic are assembled in real time, based on the specific user intent and the context of the available data.

Reply’s Prebuilt AI Apps fit into this scenario: a suite of preconfigured agentic systems that industrialise generative AI into ready-to-use solutions that can be easily integrated into the existing application ecosystem. Based on modular multi-agent architectures, curated datasets, and standard connectors to third-party sources and legacy systems such as ERPs and document platforms, Prebuilt AI Apps simplify access to information, improve decision-making quality, and increase operational efficiency.

By extending the use of AI to higher-value processes, Prebuilt AI Apps represent a concrete path from experimentation to widespread adoption, with measurable results and stable integration into core business processes. The solutions are already operational in various fields: in the insurance industry, they streamline claims management by transforming documents and reports into structured data; in procurement, they automate quote analysis and contract comparison; in HR, they enable virtual assistants and skill intelligence tools; and in marketing, they enrich campaign briefs with insights into trends, customers, and competitors.

Prebuilt AI Apps increase productivity in information-intensive functions, reduce decision-making times through intelligent document comprehension and synthesis, reduce operating costs by automating repetitive tasks, and improve the user experience thanks to intuitive, conversational interfaces. They facilitate rapid adoption, without complex infrastructure interventions, generating tangible benefits in terms of operational efficiency and reduced manual workload right from the early stages of implementation.

02.3 Silicon Shoring

The integration of AI into the software development lifecycle (SDLC) is evolving rapidly: from the tactical use of simple programming assistants to a strategic approach based on ecosystems of autonomous agents. In application factories, the role of AI is expanding from the mere execution of repetitive coding tasks to the proactive management of the various phases of the SDLC, enabling improvements in workflows, development timelines, and software quality. Reply has introduced the Silicon Shoring delivery model, which presents itself as a strategic alternative to offshoring, thanks to the efficiency of AI agents.

Silicon Shoring is available in two main configurations to adapt to different regulatory environments and the specificities of each industry sector: the “In-house” model enhances internal IT capabilities by integrating agents directly into the client company’s infrastructure; the “Managed” model offers a turnkey software factory managed entirely by Reply. This operational flexibility is useful in addressing various critical scenarios such as the rapid development of greenfield solutions, the modernisation of legacy systems, and application maintenance, which can evolve with predictive logic.

The technological engine behind this service model is Silicon Reply, a proprietary multi-agent system designed to define increasingly autonomous software factories through a suite of over sixty AI agents. Coordinated teams, made up of agents specialised in requirements analysis, development, testing, and infrastructure management, collaborate to cover the entire SDLC. Reply agents securely connect to internal knowledge sources such as code repositories and technical documentation, ensuring that each output is contextualised, secure, and compliant with industry regulations.

Reply’s experience shows that adopting these technologies requires profound change management to support IT professionals and functional staff in the new way of working and interacting with agents. Reply supports client companies on this journey, encouraging the development of skills and the maintenance of corporate knowledge bases. Process automation is simultaneously balanced by rigorous governance and an architectural oversight that ensures compliance with company standards, fostering its evolution.

AI adoption across the software development lifecycle

You indicated that your organization is adopting Al to some extent. Now, let’s look at specific phases of the SDLC. In which phases is Al currently being applied, and to what extent? (Select one per row)

Base: 405 IT Executives whose organizations either fully or partially adopted Al across SDLC

Source: Studio Forrester for Reply “From Code To Control: AI’s Takeover Of Software Development Lifecycle”

02.4 Cloud & Infrastructure

Reply supports the evolution of infrastructures and the transformation of legacy applications towards cloud-native architectures, working with leading hyperscalers: AWS, Google, Microsoft, and Oracle. The approach focuses on creating hybrid and multi-cloud infrastructures, where data can flow between different providers, ensuring resource scalability. Strategic collaboration with leading SaaS vendors (SAP, Salesforce, Adobe) rounds out the offering, enabling companies to integrate mission-critical applications into innovative environments.

Reply is specialised in managing cloud and AI cost volatility through FinOps models adopted by companies operating across various sectors. Optimisation goes beyond spending monitoring to include resource right-sizing and transaction-level cost analysis, allowing companies to accurately understand the infrastructure’s impact on each operation and allocate costs precisely. Using GreenOps also allows companies to monitor and reduce the carbon footprint of cloud services by optimising computing cycles to reduce energy consumption, combining economic efficiency with environmental responsibility.

Reply’s experience highlights a growing interest in predictive infrastructure management. Through AIOps, systems constantly monitor logs and performance, identifying early signs of malfunction before operational disruptions occur. 2025 has seen increasing use of self-healing capabilities, capable of automatically restarting services or reallocating resources in the event of an error. AI has also been used to analyse infrastructure code to prevent misconfigurations or vulnerabilities, significantly reducing the risk of data breaches.

In 2025, Reply supported several companies in integrating edge AI and edge computing solutions, bringing computing power directly to where the data is used, such as warehouses, factories, or retail outlets. This computing architecture has proven essential for low-latency applications, such as robotic control, smart camera management, or real-time quality inspection. The central cloud maintains the coordination and analysis role, while the edge ensures the immediate execution of field operations, enabling resilient, scalable, and high-performance systems.

02.5 Security & AI Governance

The adoption of artificial intelligence represents a strategic pillar for cybersecurity, enabling the shift from a predominantly reactive defence to a proactive posture. Reply integrates advanced machine learning and analytics algorithms into digital infrastructures to identify vulnerabilities, anomalous behaviour, and attack patterns in real time. Thanks to this ability to process and correlate large volumes of data and network traffic, it is possible to detect complex threats and zero-day attacks that elude traditional security systems. This approach strengthens the protection of enterprise systems, significantly improving detection and response times to security incidents.

Alongside “AI for Security” as a defence tool, in 2025 Reply organically structured the “Security for AI” domain, focusing on governance, risk management, compliance, and protection of AI models, agents, and architectures. Reply experts are supporting client companies in adapting to new regulatory frameworks such as the European AI Act and in adopting frameworks and controls that ensure safe, reliable, transparent, and ethical AI systems throughout their lifecycle.

In this context, AI red teaming plays a central role in testing the robustness of algorithms against adversarial attacks, data manipulation, and model drift risks. These activities are conducted at all stages: design, development, testing, and production, to ensure that innovation does not introduce new attack surfaces. At the same time, Reply adopts a “security by design” approach, integrating security into software development processes through DevSecOps methodologies enhanced with intelligent agents. AI is also used to orchestrate automated penetration tests and large-scale static code analyses, accelerating compliance checks with security standards from the initial project stages.

In Reply’s experience, AI is also an enabling and cross-cutting element for the protection of industrial environments and connected products, helping to reduce threat detection and incident response times. Specifically, AI-based systems continuously monitor IoT and OT networks, ensuring operational continuity and preventing intrusions into critical production and logistics systems. This technological support is complemented by strategic analysis services, including geopolitical intelligence, used to assess the impact of highly volatile international scenarios on the supply chain and corporate operations.

02.6 Data & analytics

In 2025, Reply continued its commitment to supporting companies in their journey of unlocking the value of data and the adoption of artificial intelligence, accompanying them in an environment characterised by increasing technological, regulatory, and market complexity. Reply’s approach is based on integrating data into key company processes, with the aim of fostering greater awareness, decision-making reliability, and the ability to adapt over time. In this scenario, data remains a central element of digital transformation, while artificial intelligence is emerging as a structural factor in organisational and operational models.

The latest innovations in AI, including multimodal solutions, are expanding the possibilities for using corporate information and helping make business systems more flexible, intuitive, and decision-support-oriented. Over the course of 2025, Reply’s experts have progressively shifted their focus to approaches that enable responsible innovation, with a significant focus on security, information protection, and compliance.

One area of significant growth has been synthetic data, adopted by Reply to overcome limitations related to the scarcity of real data and confidentiality and compliance constraints. Synthetic data allows AI models to be trained in secure environments, simulating complex scenarios or rare events not present in historical datasets. This approach helps reduce algorithm bias, improves the fairness and representativeness of models, and avoids the exposure of sensitive information. The use of artificial datasets also allows companies to experiment with new features and assess risk scenarios with greater flexibility and speed, significantly reducing costs and data acquisition times.

Reply’s experience in data and analytics now represents a long-term strategic infrastructure, comparable to a core industrial technology, designed to support the evolution of business models and digital services. The ability to govern these elements consistently and reliably is an enabling factor for the competitiveness and resilience of client companies. Reply will continue to invest in developing skills, methodologies, and technological vision, supporting clients in building solid data foundations for the progressive and informed adoption of artificial intelligence.

02.7 Omnichannel experience

The evolution of user experience is moving beyond traditional web and mobile interfaces to increasingly focus on conversational systems that redefine the ways brands and users interact. Reply supports this evolution by combining service design, technology architectures and process consulting to build scalable infrastructures, capable of supporting dynamic business models. This strategy provides companies with the flexibility needed to respond to market fluctuations while ensuring rigorous service and brand consistency throughout the entire customer journey, from discovery to loyalty. Service personalisation is also enhanced by AI, which supports real-time data analysis to drive recommendation engines capable of instantly adapting promotions to the user’s context.

At the production level, generative AI is radically changing the content value chain, optimising marketing processes through the automation of creative production, with significant potential for scaling the production of increasingly personalised digital assets. The integration of advanced text-to-video and text-to-image models into structured workflows allows for the generation of high-resolution multimedia variants, drastically reducing the time-to-market of global campaigns and reducing production costs.

2025 also saw the rise of generative UI, which allows the entire user interface to be generated in real time: instead of forcing content into predefined interfaces, AI dynamically builds interactive elements, maps, simulations, and rich formatting in response to the user’s intent. This capability transforms the output of AI models and platforms into tailored visual experiences, offering a mode of interaction that is superior to standard formats in terms of immediacy and contextual relevance.

The convergence of physical and virtual contributes to the evolution of the omnichannel experience, leveraging spatial computing and immersive technologies. Reply develops innovative pipelines for the production of three-dimensional assets by combining full-body motion capture with applied generative models. One area of convergence between these three-dimensional capabilities and artificial intelligence is that of digital humans, widely used both as a conversational interface for accessing corporate knowledge and training, and in customer interaction during pre-sales and post-sales support.

Digital Experience: trends in 2026

PLATEAUING TRENDS
Trends that experienced rapid
growth throughout, but have now
reached a point where expert
interest has levelled off

ENDURING TRENDS
Trends with continued and
accelerated expert interest
throughout the last 12 months

DECLINING TRENDS
Trends with decreasing
expert interest throughout
the last 12 months

ACCELERATING TRENDS
Trends that are gaining traction
and showing signs of rapid
expert interest growth
throughout the last 6 months

Source: Reply study “Digital Experience Trends 2026” - Reply analysis on Sonar Reply data

02.8 Physical AI

AI applied to the physical world is opening a new phase in the evolution of spatial intelligence, namely the ability of systems to understand and interact with the surrounding environment. In this context, “world models” represent the foundational architecture of intelligent physical devices, enabling them to learn the laws of physics and predict the outcomes of actions without relying on preconfigured static maps. At the core of these systems is computer vision, which transforms raw sensory inputs from high-resolution cameras, LiDAR and tactile sensors into a true spatial understanding. Advanced deep learning algorithms then enable real-time object recognition and scene interpretation, providing the perceptual foundation required for autonomous decision-making in complex and unstructured scenarios typical of urban and industrial environments.

In industrial and logistics contexts, Physical AI is being applied through the use of autonomous mobile robots (AMRs), inspection drones, and collaborative robots (cobots) designed to optimise production flows. These systems automate high-risk tasks such as monitoring critical infrastructure or operations in hostile environments, increasing operator safety and operational resilience. The integration of intelligent sensors and visual inspection systems also fuels predictive maintenance models that reduce downtime and optimise goods movement, ensuring high levels of reliability and business continuity through constant monitoring and rapid response to unexpected events.

AI also supports the adoption of digital twins and high-fidelity simulated environments where algorithms are trained to perform complex tasks before being deployed on physical hardware, including humanoid robotic systems. These simulations generate synthetic data that replicate real-world conditions, including rare or dangerous edge cases that cannot be physically captured, bridging the gap between simulation and reality. In the short term, teleoperation serves as a crucial bridge to full autonomy, allowing human operators to guide robotic systems through complex tasks and generating high-quality demonstration data for training.

In 2025, the growing integration of AI has further expanded the offering of connected products and services. Reply has supported companies in developing AI-powered IoT ecosystems, ranging from smart homes to digital healthcare, to smart vehicles equipped with virtual assistants and predictive diagnostic capabilities. The integration of natural language processing technologies now enables interaction via intuitive voice interfaces, improving the user experience and contributing to energy efficiency optimisation. Connectivity enables advanced servitisation models, transforming physical products into services managed and monitored remotely throughout their lifecycle.

02.9 A look ahead to 2026

In 2026, Reply will continue to support client companies’ innovation by leveraging the rapid maturation of AI, thanks to international and multidisciplinary research and development groups. Significant attention will be paid to the transition from conversational systems operated directly by the user to agentic systems to which tasks can be delegated to be performed autonomously and subsequently evaluated by humans. 24/7 agentic loops will be a key factor in the era of agentic AI: Reply will continue to support the redesign of business processes around this shift, from linear workflows to “judgment cycles” (drafting, critique, verification, decision).

In parallel, Reply is developing its approach to Responsible AI through a developer-focused toolkit, including the creation of automated bias-checking systems, fairness assessment agents, and risk-rating bots, aligned with the European AI Act and OECD guidelines. Reply teams are addressing AI sovereignty as a critical architectural decision, developing reference architectures for sovereign clouds and local inference. This ensures that client companies can implement advanced AI capabilities while maintaining complete control over where their data resides, who manages the models, and what jurisdictions they are subject to.

In this context, Reply is exploring the use of small language models (SLMs), compact and highly efficient models. SLMs could become the key to democratising AI inference, enabling rapid real-time processing locally, without relying on the cloud. This approach is crucial for industries with strict data sovereignty requirements or connectivity constraints, such as healthcare, manufacturing, and defence. By processing sensitive data locally, this approach can reduce inference costs and privacy risks.

As classical computing approaches its physical limits for specific optimisation and simulation workloads, Reply is actively validating hybrid quantum-HPC architectures, focusing on use cases such as quantum machine learning applied to time series analysis for financial risk modelling and complex logistics optimisation. Reply is also helping companies prepare for post-quantum cryptography to protect corporate data and digital assets from potential decryption threats enabled by the combination of advanced AI and quantum technologies.

03.1 Axulus Reply

Axulus Reply is a generative AI-based engineering platform for digitalisation processes in industrial contexts. Through AI models and computer vision, it supports companies in the manufacturing, infrastructure and logistics sectors in improving the effectiveness of production and operational processes. The platform provides ready-to-use frameworks and modular digital workflows, enabling progressive and scalable adoption. Solutions based on Axulus Reply are used in airport and industrial environments to support the optimisation of operations, as well as the safety and quality of services and products.

03.2 Brick Reply

Brick Reply is Reply’s digital “as a Service” platform supporting the transformation of industrial production processes. Based on a microservices architecture and an open ecosystem, Brick Reply ensures end-to-end management of industrial activities and full integration with enterprise and factory systems through standard APIs. In the course of 2025, a cognitive engine was integrated, forming the foundation for agents specialised in the main MOM domains, including production, planning, quality, performance and materials management. This architecture makes it possible to analyse large volumes of heterogeneous data and to support operators and plant managers through advanced analytics and operational recommendations.

03.3 Discovery Reply

Discovery Reply enables the centralised management of digital content, multimedia assets and product data. Based on a modular and omnichannel model, the platform ensures governance, versioning and consistent distribution of content across different touchpoints. In 2025, the introduction of agentic AI extended the platform’s capabilities towards a content intelligence model, with specialised agents for the creation, search, processing and transformation of content, also accessible through a conversational interface. The evolution has also involved semantic search, with retrieval mechanisms based on meaning, context and relationships between content.

03.4 LEA Reply

LEA Reply is Reply’s platform for orchestrating supply chain execution processes in complex and automated logistics environments. Based on a modular microservices architecture, it covers activities such as warehouse management, inventory, distribution and delivery of goods, integrating with IT systems, automation, robotics and IoT. Through the integration of AI, advanced analytics and computer vision, LEA Reply enables end-to-end visibility of logistics flows and supports proactive management of operations. In 2025, the evolution of LEA Reply extended its ability to orchestrate heterogeneous logistics ecosystems, integrating e-commerce scenarios, omnichannel models and drop shipping within a single operational framework. The introduction of GaliLEA, an AI assistant based on a multi-agent architecture, has further enhanced decision-making support, allowing users to interact with the platform in natural language and analyse operational and business metrics in a more immediate and contextualised way.

03.5 Neurons Reply

Neurons Reply is a platform for the creation, management and orchestration of agentic AI solutions in enterprise environments. Designed to foster the adoption of artificial intelligence within organisations, it enables users with different levels of expertise to aggregate, create and customise intelligent agents to support business processes, automation and internal collaboration, promoting a culture of collective intelligence and co-creation. Built on a modular and model-agnostic architecture, Neurons Reply provides access to a variety of language models and is compatible with interoperability standards such as A2A (Agent-to-Agent) and MCP (Model Context Protocol) for structured access to tools and data, as well as for developing advanced interaction approaches such as MCP-UI. Thanks to a secure and compliance-ready framework, Neurons Reply facilitates the scalable adoption of agentic AI in enterprise contexts, improving productivity and knowledge sharing.

03.6 TamTamy Reply

TamTamy Reply, born as an enterprise social network platform to improve internal communication and collaboration, has evolved into a complete digital ecosystem that integrates knowledge management, learning experience (LXP) and content distribution. In 2025, Reply applied AI capabilities to its Enterprise Social Network, Learning Management System and Digital Experience Platform (DXP) modules, integrating a proprietary platform of intelligent agents to automate processes, manage information and support content distribution and learning pathways. The DXP module enables the creation of dynamic and personalised mini websites for internal communication, event management and content distribution, while the Learning area provides adaptive, AI-supported learning pathways tailored to the needs of corporate users.

03.7 Ticuro Reply

Ticuro Reply, certified as a Class IIa (CE) medical device, is designed to support prevention and continuity of remote care. The platform uses AI and IoMT (Internet of Medical Things) technologies to integrate with medical devices and wearable sensors, facilitating collaboration between patients, caregivers and healthcare professionals throughout the care pathway. It also provides fast and remote access to clinical information from heterogeneous sources, offering a complete and up-to-date view of the patient. In 2025, it was tested in the Italian market as a telemedicine platform across six regions, demonstrating reliability, scalability and compliance with regulatory requirements.

03.8 X-RAIS Reply

The X-RAIS Reply platform applies artificial intelligence to the medical field and is designed to support radiological diagnostic processes through deep learning techniques. Specialised in various diagnostic methods and specific anatomical regions, it uses advanced image recognition technologies to assist radiologists in identifying anomalies and pathologies, helping to improve the accuracy and timeliness of diagnoses. Training of algorithms for the identification and classification of microcalcifications in mammograms and breast density was completed in 2025.

04.REPLY LABS

Reply Labs are spaces for innovation and co-creation where new technologies are applied to real-world contexts, generating solutions that can be integrated into everyday activities and enable new business models.

04.1 Area42

Area42 is dedicated to the exploration and testing of robotic solutions such as quadrupeds, rovers, drones, and interactive humanoid robots. Reply experts conduct tests and experiments on the algorithms that govern the perception, mobility, and manipulation of robots. They perform tests and experiments on reducing AI models so they can run on edge computing. They implement edge-to-cloud continuum solutions to allow software controlling mobile devices to access the correct network resources when needed. They also generate synthetic data for model fine-tuning. The most innovative technologies are validated with real-world experiments, thanks in part to collaboration with universities and international research centres.

04.2 Area Phi

Area Phi supports client companies in translating technological innovation into concrete strategies for their sectors. Key activities include: developing empathetic digital humans, integrating conversational AI and generative AI; exploring immersive technologies such as extended reality, gamification, and AI to create engaging experiences, including digital escape rooms, immersive e-learning, and interactive games. In the field of edge computing, the lab develops solutions for manufacturing and renewable energy, improving efficiency through real-time analytics, predictive maintenance, and defect detection with AI.

04.3 Cybersecurity Lab

The Cybersecurity Lab provides companies with an advanced environment for analysing, evaluating, and simulating security scenarios in areas such as cloud computing, secure software development, application and data protection, and network infrastructure. The Cybersecurity Lab includes demo units dedicated to attack tests and simulations, threat modelling, and security analyses of hardware and software components. Its activities extend to industrial security, IoT, automotive security, and smart building security, allowing client companies to effectively evaluate and strengthen their defence strategies.

04.4 IoT Validation Lab

The IoT Validation Lab is dedicated to the design, integration, validation, and implementation of IoT solutions and connected products, with a significant focus on environmental sustainability and energy efficiency. Thanks to end-to-end expertise and advanced technological instrumentation, Reply experts support client companies throughout the entire lifecycle of IoT solutions, offering consulting services, connectivity testing, and device pre-certification activities. The IoT Validation Lab develops and analyses applications in the automotive, telecommunications, manufacturing, energy, and logistics sectors, enabling the sustainability of IoT infrastructures to be assessed through analysis of material degradation and component lifespan.

04.5 Test Automation Center

The Test Automation Center ensures continuous quality monitoring of strategic products and services for the business, using a structured framework and proprietary validation methodologies. The goal is to support client companies in managing the entire lifecycle of products and services, both during the development phase and after market release, facilitating the early identification of potential critical issues and the timely definition of corrective actions. The integration of AI into quality assurance tools increases the efficiency of key processes, including test selection, data preparation, and maintenance activities, ensuring high levels of reliability even in technologically complex environments.

05.ALLIANCES

Over the years, Reply has built and consolidated a broad ecosystem of collaborations with innovative companies, developing strategic relationships with leading technology vendors to design and implement highly customised solutions aligned with each client’s specific business needs.

05.1 Adobe

Adobe and Reply collaborate on solutions for unified customer experience management and the orchestration of B2B go-to-market strategies. As an Adobe Platinum Partner, Reply achieved seven specialisations in 2025 on Adobe Experience Manager and Adobe Commerce, delivering solutions that cover the entire Adobe Experience Cloud platform. During the year, it also received the “2025 Adobe Experience Manager Rockstar” award and was recognised by leading market analysts for its expertise in Adobe technologies and the use of AI to optimise clients’ content supply chains.

05.2 AWS

Reply is an AWS Premier Tier Service Partner with extensive experience in migrating enterprise systems to the AWS cloud. It offers a comprehensive portfolio of services, including content processing and distribution solutions, development and integration of custom enterprise applications, as well as end-to-end support with 24/7 maintenance and operational management services. Reply has certified expertise in numerous strategic areas, including Agentic AI, Data & Analytics, DevOps, Oracle, Migration, Internet of Things, Industrial Software, SaaS, Machine Learning, Financial Services, Security, Retail, Energy, and Automotive. In 2025, AWS recognised Reply as the EMEA Energy and Utilities Partner of the Year and Italy Partner of the Year. Reply has also signed a Strategic Collaboration Agreement with AWS for the development of solutions based on generative AI. Reply was also named an Implementation Partner for the launch of Amazon Bedrock AgentCore, a platform dedicated to the development and operation of scalable AI agents.

05.3 Google

In 2025, Reply strengthened its collaboration with Google Cloud and Google Ads in Europe, the United Kingdom and the United States. As a Premier Partner and Managed Services Provider of Google Cloud, it supported clients in adopting Gemini Enterprise for the creation and deployment of AI-based agents and workflows in controlled environments. During the year, Google Cloud recognised Reply as 2025 Partner of the Year for the UK and Ireland. At the same time, Reply’s creative agencies enhanced their expertise in Google Ads and AI content production, delivering digital campaigns for clients with video content created using the Google Nano Banana Pro and Veo models.

05.4 Microsoft

Reply is a global Microsoft partner and a leader in developing solutions across the three main areas of the Microsoft ecosystem: AI Business Solutions, Cloud & AI Platforms, and Security. In Europe, the United Kingdom, the United States and Brazil, Reply operates through a network of specialised companies covering all Microsoft Solutions Partner Designations, with vertical expertise across various industrial sectors. With deep knowledge of Microsoft AI and Copilot technologies, Reply delivers innovative solutions applied to multiple business areas. In 2025, Reply received the Partner of the Year awards for Global Social Impact and Inclusion Change Maker, as well as Emerging System Integrator of the Year for the USA, and was selected for a dedicated adoption programme for large clients in the EMEA region.

05.5 Oracle

Reply operates as an Oracle Cloud Service Provider, supporting clients with managed services and the implementation of solutions on Oracle Cloud Infrastructure. In 2025, it received the EMEA Cloud Service Provider Partner Award. In the same year, Reply expanded the use of Oracle solutions for Finance, Supply Chain, Planning and Production, delivering international projects based on ERP Cloud and NetSuite, particularly in the manufacturing and healthcare sectors. The company also achieved seven Service Expertise recognitions and broadened its offerings in the public sector and with international institutions.

05.6 Salesforce

Reply is a certified Salesforce Consulting Partner with a presence in Europe and the United States. It supports clients across the main cloud areas of the Salesforce ecosystem, including Sales, Service, Marketing and B2B and B2C Commerce. Its offering also includes integration with MuleSoft, analytics with Tableau, digital collaboration through Slack, and data integration and governance solutions based on Informatica. Reply also works on Salesforce Industry Clouds and advanced platform components, including Field Service, Account Engagement, Data 360, and AI-based solutions such as Einstein and Agentforce. In 2025, it was selected as a launch partner for Salesforce Retail Cloud in the EMEA region, developed specialised expertise in the automotive sector, and delivered one of the first implementations of Financial Services Cloud in Europe.

05.7 SAP

Reply’s SAP expertise spans a cloud solutions portfolio that includes SAP S/4HANA Cloud, delivered through RISE with SAP and GROW with SAP, Business Process Transformation, SAP Customer Experience, SAP Business Technology Platform, SAP Business Data Cloud, SAP Digital Supply Chain, SAP Ariba, SAP Concur and SAP Business AI. In 2025, Reply strengthened its role as an SAP Lighthouse Partner for SAP Business AI and participated in the SAP Joule Early Adopter Care programme. During the same period, it expanded its capabilities in multi-agent frameworks based on SAP Business AI, achieved the GROW with SAP designation, and won the SAP Quality Awards for the twelfth consecutive year.