After the initial wave of generative AI, the whole world now seems to be talking about agents

And while the words agent and agentic are now suddenly everywhere, they have a long history and well-established meanings.

Defining what you need is important

Whilst it might seem simplistic to focus on definitions, it is crucial to do so for this topic. Why? Because there are many different definitions of agents and agent systems.  If we are ever to form a coherent, interconnected ecosystem of agents, we need to start with clarity and consistency around the terminology. Ironically, the terminology around agents is some of the oldest and best-established in the AI field. It is also interesting to consider people’s backgrounds and interests. People who sit more on the business side tend to define an agent based on how it affects the business whereas technical people tend to define an agent based on what it can functionally do. Here, we define key concepts to help demystify this topic for all. 

An agent

An agent is any entity that works on behalf of another entity, working to accomplish high-level objectives often using specialist capabilities. Agents have the autonomy and authority to take actions that modify their world.

A key aspect of this definition is the ability to take action – literally exercising agency. AI can be a great assistant, but if it does not have the ability to take action, it cannot be an agent. An agent, however, does not necessarily have to use AI. Many non-AI systems are agents too (a smart thermostat is a simple non-AI agent). Similarly, not all AI systems are agents. So, let’s take a deeper look into what some of the key terms are around agentic systems to better understand what an agent is and is not.

. 

Autonomy, authority, and agency 

IThese three terms could well be thought of as quite similar, however it is important to define them and discuss how they differ from each other and why it matters in this context.

• Autonomy is a measure of the degree to which an entity can independently make choices.
• Agency refers to the degree to which an entity has the capacity to act on those choices.
• Authority refers to the specific scope or limitations of the actions an entity can take.

All of these are continuous spectrums, not binary properties. The thermostat mentioned earlier has a high degree of autonomy (it can decide what needs doing without human intervention), a high degree of agency (it can take those actions without oversight), but a low degree of authority (it can only do one thing – switch the heating on or off). Using this example, you can imagine many agents and debate how much autonomy, agency, and authority they have.

One more thing to consider here is that the designers of agentic system are the ones deciding how much autonomy, agency, and authority agents have. So, we should all ask these questions when making agentic systems to ensure they can operate under the conditions they need but do not have complete freedom beyond the need for the use case. These three properties provide a useful framework to discuss and assess the risks and opportunities of agents. An agent with high autonomy, agency, and authority could be extremely powerful, but also very risky. Therefore, what is the necessary level of autonomy, agency and authority? It is up to you, as the human designing the system, to decide!

A multi-agent system (MAS) 

So far, we have talked about what an agent is, but not much about what happens when you have lots of agents working together. Whilst there are lots of great use cases that require single agents, often the value will be had with a multi-agent system. But what is a multi-agent system (MAS)? Simply, we define a MAS as a system made up of multiple independent agents that operate in the same environment.

It is also worth noting that systems that use agents are sometimes called agentic architectures/ frameworks. Now that we have defined many of the key terms, let’s move on to some of the key concepts.

World models  

Agents operate within a specific “world”, representing the totality of what they can sense and act upon. This could be a narrowly defined software environment, or the actual, physical world. Coming back to our thermostat example, with a limited world model, the thermostat only knows about temperature. An advanced thermostat with a richer world model might understand occupancy patterns, weather forecasts, utility pricing, and user preferences.

This comprehensive understanding allows it to make decisions that appear intelligent rather than merely reactive – turning down heating before you leave or pre-warming before expected return – building trust through apparent understanding of context. An overly simplistic world model can lead to poor performance. If a customer service agent does not have good contextual information about the customer and their situation, its advice would likely be of very low quality. World models are something that all humans have, and while they may differ slightly between people, our shared model of the world allows humans to collaborate, anticipate, and empathize with each other in order to solve tasks efficiently.

World models are essential for AI to be able to be trusted. They allow us to understand whether the AI’s success or failure was due to the right reasons, and not simply because there was a misalignment between us as humans giving it instructions and its understanding of the environment it was operating in. 

Relationship between agents and LLMs 

We previously said that agents do not need to have AI to meet the above definition of an agent. This can be extrapolated further to say that AI agents do not need to have an LLM core. Agentic and multi-agent systems may not include any Gen AI at all. This is easy to see using our previous definitions of autonomy and agency: clearly LLMs are not required to enable either of these concepts.

It might seem obvious to say agents don’t have to be LLMs, but most of the examples of agents that people mention today do have an LLM core. It is also worth mentioning that often things with an LLM core are called agents, but do not have the ability to exercise agency at all. As a result, these would not meet our criteria for an agent.

Agents and multi-agent systems have been a cornerstone for AI for well over 30 years. The reason why agentic architectures have taken off in the LLM era is because LLMs provide a rich and natural way for humans to communicate goals with AI systems, and natural language provides a way for agents to communicate with each other. The classic phrase of human language being the way to communicate about anything in an inefficient and imperfect way, rings true here. , rings true here.

Five additional dimensions of multi-agent systems (MAS) 

We can now look a little deeper into what a multi-agent system is and how we can classify it. On one hand, we can describe agents and their properties of autonomy, agency, and authority. On the other, we can describe dimensions of the whole system.

Here, we propose five dimensions that help us better understand multi-agent systems. The first dimension is size. Then, we talk about heterogeneity. While homogeneous systems are those where agents share similar roles (often called swarms), heterogeneous systems feature specialized agents that handle complex tasks. Heterogeneous systems can self-organize and coordinate to solve a problem but require sophisticated coordination. We then consider the concept of centralization. Centralized systems require rigid structures and orchestration, but are more controllable and explainable. Decentralized systems distribute decision-making broadly, enhancing scalability and resilience, but complicating system coherence and control. These three dimensions may seem like the more the merrier, but larger, heterogeneous, and decentralized systems are harder to control.

Now let´s go back a little to describing aspect of agents rather than the system with specialization. Generic agents often exhibit greater autonomy, capable of flexible decision-making in diverse scenarios, but are rarely able to complete complex tasks. Specialized agents, whilst highly skilled in specific domains, typically exhibit higher agency and lower authority, executing only narrowly defined tasks. The reason why a dimension at an agent level is in this section is because while agentic systems often have a mix of specialization of agents, the system itself can also be described in terms of specialization too.

Lastly, there is the degree to which the system is deterministic or not. Determinism describes how rigid and predictable a system is. Basically, if you do the same thing multiple times, a deterministic system provides the same answer every time. This is where we are seeing lots of change with the wave of Gen AI. Typically, agentic AI systems have been very deterministic. If the thermometer detects a level at 20°C, then it will turn the heating off. Therefore, a fully deterministic system will always produce the same outputs given the same inputs. Their performance will always be the same, which is both good and bad. By contrast, non-deterministic systems might adapt and change their behavior over time. This allows them to improve over time, but also runs the risk that their behavior might become worse or even harmful. It is therefore important to understand how to manage this emergent behavior and monitor it to ensure the desired emergent behavior is obtained.

These dimensions interact in intricate ways, and understanding them is key to designing multi-agent systems that have the desired performance and trustworthiness across diverse architectures and use cases.

To learn more about these topics and explore them further, visit Robert Engels’ blog here. 

Maturity model for autonomous AI enablement 

We have spoken a lot about what agents are and what an agentic AI system is, but how can we understand them better? Understanding the degree of agency and autonomy in a system is vital to understanding both its power and its risk profile. For example, if we take well-known agents that are used by humans today such as real estate agents, travel agents, and insurance agents, we can plot how much autonomy and agency we give them. We can also understand why we would not want to give full autonomy or agency to them.

We want a real estate agent to be autonomous in selling our homes, but to not have the agency to agree to the sale price without us. We might give travel agents agency within relatively tight constraints to make bookings on our behalf, but not make major changes to dates or destinations. We would want insurance agents to have reasonably high autonomy and agency; they can take out insurance, make sure we have the coverage we need, and we trust they are more competent at that than us.

If we look in the extremes, we find high agency with low autonomy such as sports agents. They negotiate contracts and agree terms, but only when the athlete gives them permission to talk to someone. An extreme example of high agency with high autonomy is a secret agent. Here, the mission is provided, but the agent can decide entirely how they complete it and have full agency, even beyond the law, to act however they choose to achieve the outcome. Hopefully this section helps you realize that the level of autonomy and agency we give to human agents in our world today is the result of the decisions we make and the desired outcomes we have in mind. We must think of autonomous agentic AI systems with the same clarity. These systems will perform within the bounds we give them and optimize against the purpose we assign to them. 

A complex landscape 

Whilst the mainstream narrative talks about agent implementations as a simple architectural pattern, our exploration of the many attributes and dimensions of agents shows that this is a much deeper topic. Autonomous systems and AI agents will be a defining feature of the technological landscape of our future, and understanding the qualities and dimensions of agency will help us navigate this complex and exciting future with confidence

About the Capgemini Group AI Lab

We are the AI Futures Lab, expert partners that help you confidently visualize and pursue a better, sustainable, and trusted AI-enabled future. We do so by understanding, pre-empting, and harnessing emerging trends and technologies to ultimately make trustworthy and reliable AI that triggers your imagination, enhances your productivity, and increases your efficiency. We will support you with the business challenges you’re currently aware of and the emerging ones you will need to be aware of to succeed in the future.   We create blogs, like this one, Points of View (POVs), and demos around these focus areas to start a conversation about how AI will impact us in the future. For more information on the AI Lab and more of the work we have done, visit this page: AI Lab. 

Today’s industrial challenges are rarely isolated. They’re interconnected. Productivity, sustainability, digital transformation – they’re all part of the same conversation.

But too often, organizations are forced to tackle them separately. At Hannover Messe 2025, we at Capgemini will be showing what happens when you take a different approach.

In partnership with Microsoft, we’ll be bringing intelligent industry to life through real-world demonstrations, industry-led conversations, and practical examples of how collaborative thinking can drive better results. It’s a chance to see how advanced AI, digital twins, and real-time data can drive real progress across manufacturing, automotive, defense, and electric battery innovation.

Come and see it in action

At the center of our stand is the Digital Twin Cockpit – an interactive experience for production engineering which allows faster ramp-ups and better production design validation while providing complete data scalability and integration for operations use-cases. The Digital Twin Cockpit integrates a Unity viewer with Microsoft technologies including GenAI copilot for intuitive querying of digital twin information.

It’s fully interoperable, able to plug into different CAD and data sources, and brought to life through a VR-enabled 2D interface – giving you an immersive, hands-on look at what’s possible when digital and physical realities come together.

Our experts Olivier Saignes and Nicolas Vasseur will also be sharing insights about new digital shopfloor performance during a live presentation at the Microsoft booth on Thursday April 3 at 10:00. They’ll explore how Capgemini and Microsoft are working together to unlock performance, resilience, and scale in today’s digital factories – and what that means for your next step forward.

Why it matters

Capgemini brings together extensive manufacturing expertise with world-class engineering capabilities – making us a trusted partner for industrial transformation at scale. In close collaboration with Microsoft, and alongside key partners like NVIDIA or Siemens, we combine best-in-class technology with sector-specific insight and hands-on experience. It’s how we help clients move faster, think bigger, and deliver more – with a clear path to value.

The organizations that are moving fastest right now are those finding ways to connect across silos – combining data, teams, and technologies to solve overlapping challenges at once. That’s what we call Compound Solutions

It’s a joined-up way of working that helps clients make progress in several areas at the same time. Whether it’s increasing efficiency, reducing emissions, or modernizing infrastructure, the impact is greater when those goals are tackled together.

It’s more than a stand – it’s a space for real conversations

On Monday March 31 at 17:00, Microsoft’s Dayan Rodriguez and Capgemini leaders Pierre Bagnon and Lydia Aldejohann, will be speaking live on stage about:

• Major trends currently impacting the manufacturing domain
• Advancements in AI bringing intelligent manufacturing closer to reality
• Real-world success stories from intelligent manufacturing
• Proven ways to overcome industry challenges
• The power of an open, collaborative ecosystem

Our thought leadership sessions will also share practical insights on how clients are applying AI, robotics, spatial computing, and digital twin technologies. You’ll hear stories of what’s worked, what’s changing, and how to get ahead.

We’re recognized for our results

Capgemini was recently recognized by Everest for its leadership in intelligent industry – and we’re already making a real difference across some of the most advanced, high-performing sectors. From helping manufacturers scale transformation to supporting defense clients with secure, intelligent operations, we’ve built a strong track record by delivering where it counts.

Visit Capgemini at Hannover Messe

If you’re attending Hannover Messe, we’d love to see you.

Come by the Capgemini booth to:

• Try the Digital Twin Cockpit demo for yourself
• See how Agentic AI is already transforming industrial operations
• Talk to our experts about the opportunities for your organization
• Explore what Compound Solutions could mean for your business

We’ll also be livestreaming some of our sessions if you are unable to attend in person.

We’re ready to show you what’s possible when industry meets impact. Where innovation becomes action. Where you get the future you want.

There’s a crucial need to demystify bio-engineering solutions to demonstrate their value, gain public trust and understanding, and drive adoption and economic growth.

The COVID-19 pandemic was a moment in history, when global governments, private institutions, academia, and society as a whole, coalesced around the need and to benefit global wellbeing. When a vaccine was produced and rapidly approved through extraordinary measures, the collective relief was palpable, but there was an unexpected problem.

Public trust in science, representation, and a lack of bio-literacy around mRNA vaccine technology challenged the progress of the global vaccination program. Governments worldwide invested hundreds of millions in fighting disinformation about bio-engineering solutions.

For a new innovative solution to be adopted, the people it serves must trust that it will work in their best interests, and trust that those developing solutions have their best interests at heart. In the case of such new and emerging deep-tech as bio-engineering solutions, with their implications to natural evolution, building trust and understanding is more important than ever.

There are clear steps we can take to build this trust, to drive adoption of bio-engineering solutions and related socio-economic growth.

Low levels of literacy and trust: policy makers, citizens, and markets must better understand bio-engineering solutions to unlock their potential

Recent data from the Capgemini Research Institute highlighted the growing divide between industry innovation and public understanding: 96% of companies reported active engagement with, or plans to develop, bio-solutions. And yet, research by the UK Department for Science, Innovation and Technology (DSIT) showed that 76% of UK respondents feel they lack sufficient information to make informed decisions about bio-based products. The same report found that 61% have never heard of engineering biology. Although these findings are UK-specific, similar trends are evident globally. For example, the CRI report also revealed that 65% of startups and corporations view low public bio-literacy as a significant barrier to adoption.

The consequences of this gap are profound. One recent study showed that decades-old public resistance to the distinct domain of genetically modified organisms (GMOs) shaped overly cautious regulatory frameworks that impede advancements in plant genome editing to this day. As a result, solutions to critical challenges, such as food security, continued to be delayed. More recently, hesitancy around mRNA vaccines has shown how a lack of communication can erode trust and create a vacuum for misinformation to grow. Without proactive public engagement, societally-valuable bio-based innovations such as bio-based plastics, precision farming, and sustainable fuels risk similar skepticism, potentially stalling their adoption. Governments and industries must act collaboratively, ensuring consumers feel like active participants in this journey, not passive spectators.

Business and government: shared responsibility to unlock economic growth through engineering biology vision

The challenge of bridging the gap between innovation in bio-engineering solutions and public understanding is a shared responsibility. It demands coordinated action from both businesses in industry and policymakers.

Businesses must go beyond simply creating innovative bio-products – they should actively engage citizens and a broad stakeholder group throughout the development process by employing user-centered and co-design techniques. This not only builds trust and understanding with consumers but of course ensures stronger market fit and a greater chance of commercial success.

To unlock the economic growth potential of bio-engineering solutions, policy makers also need to engage in similar user-centered and co-design techniques early in the policy making process through techniques like ‘Policy Labs.’ Policy Labs help rapidly map ‘policy whitespace’ where a new emergent theme or technology means there are few existing policy ideas or policies to build on. If policy teams become blocked due to a lack of understanding of complex new technologies, this prevents private markets from progressing due to a lack of support or legislation and regulation that unnecessarily blocks development. 

The importance of Policy Labs

Policy Labs are a proven technique to test early policy hypothesis, to ensure policy solutions will work for all stakeholders. For instance, if there’s to be a proposed investment by the Department of Environment Farming and Rural Affairs (DEFRA) into alternative proteins or GM crops, the farming industry should be engaged to test early policy hypothesis to see whether proposed solutions, services, regulations or legislations are likely to empower farmers with new economic means. Or disempower them by creating new competition in the markets or an even more complex post-Brexit funding landscape.

A Policy Labs approach also ensures better cross-departmental working and inputs from the relevant government bodies that should be involved, across such a broad problem space. For example, in the UK, the Department of Business and Trade (DBT) or Department of Science Innovation and Technology (DSIT) might work with DEFRA to provide a better holistic approach and collaborative understanding to launching new solutions by data sharing agreements. Subject Matter Experts (SMEs) and private sector specialists in Engineering Biology could also be engaged through the policy labs process to ensure policy ideas are technologically viable, operationally feasible, and economically scalable in markets.

Open and honest dialogue

Governments should also be creating an ongoing dialogue with citizens to ask what assurances they need to feel confident about bio-engineering solutions (products and services) and how perceived risks can be addressed. Transparent, accessible communication mechanisms are key, alongside independent oversight to ensure safety and accountability. As the Engineering Biology Research Consortium (EBRC) puts it: ‘Public engagement, improvement of public perception, and building trust are critical factors for the growth of the bio-economy and market.’

A call to action for bio-economy solutions

Engineering biology has the potential to solve some of the world’s most pressing population scale challenges, from climate change to healthcare. But without better collaboration between governments, industry, and the markets they serve, progress on bio-engineering solutions will stall. By employing co-design both in the product development and policy making lifecycles, we will build better understanding and trust among stakeholders and citizens, while enabling policymakers and businesses to ensure bio-solutions not only innovate but also meet public needs.

At Capgemini, we provide end-to-end support in Engineering Biology and AI for Science strategy, helping public and private sector clients accelerate bio-solutions while addressing the challenges outlined here. From fostering public bio-literacy and engaging citizens to building transparency into innovation and driving growth through regulatory collaboration, we leverage our insight drawn from practical laboratory bio-engineering, to advise our clients on delivering impactful, trusted, and scalable outcomes. Together, we can draw up a bio-economy strategy that maximizes public trust in the solutions that will shape the world for years to come.

The time is now to act. Together, we can do more than just create bio-solutions; we can create a bio-ready society. Get in touch to explore how we can help accelerate your engineering biology journey.

Capgemini’s Job Fair solution improves CV processing and communication by making the recruitment process more efficient and transparent for candidates.

At one time or another we have all felt the frustration of sending out dozens of job applications, each tailored to a specific job or company, only to receive no response from any HR team whatsoever.

This lack of communication can leave candidates wondering if their application was even considered or how they could improve for future opportunities.

Simplifying CV processing

Capgemini saw overcoming this problem as a challenge which is why we set out to improve our CV processing capabilities, helping us deliver an intelligent and connected “consumer-grade” people experience to any potential candidate who engages with us.

Our Job Fair solution ensures timely communication with candidates, including those who will not move forward in the recruitment process.

Our process allows HR teams to quickly inform candidates when their recruitment journey has concluded without an offer. These messages are phrased positively to motivate candidates for future efforts while maintaining transparency and efficiency.

This approach helps Capgemini build a reputation as a company that cares about career growth, even for those who haven’t worked with us.

Tackling CV processing challenges with precision

Capgemini’s Job Fair solution uses Optical Character Recognition (OCR) and Microsoft’s Power Automate technology to extract key information such as email addresses, and phone numbers from various documents quickly.

Furthermore, our Job Fair solution’s simplicity and flexibility, combined with its straightforward interface, requires minimal training to operate effectively. This ensures improved CV processing comes with minimal disruption, and enables HR teams to meet a wide range of business needs without developing a new, expensive solution.

Delivering award-winning HR processes

We understand that companies need to focus on bringing their people, processes, and technology together to deal with whatever their business might face, moving them closer to becoming a truly Connected Enterprise.

This mindset is why Capgemini recently won a Gold Medal in Brandon Hall’s Excellence in Technology Awards, 2024. This highlights Capgemini’s commitment to implementing effective, easy-to-use applications that address its clients’ needs at speed, leaving a positive impact on their businesses.

But that’s not all. Capgemini also won a Silver Award in Brandon Hall’s HCM program, 2023, which clearly demonstrates that Capgemini is among an elite group of exceptional HR service providers.

To discover more about how Capgemini’s Intelligent People Operations put your employees at the heart of HR operations, across your talent acquisition, HR administration, payroll, and HR analytics functions, to deliver strong and sustainable business value, contact: alicja.watorek@capgemini.com

The struggle to meet rising customer expectations

Customer expectations for financial services firms have never been higher. Whether submitting a mortgage application to a bank or contacting an insurer to file a claim, consumers demand seamless, hyper-personalized, and efficient service at every interaction. However, many financial services contact centers still rely on outdated models that struggle to meet these expectations. Long wait times, fragmented communication channels, and manual processes create frustrating experiences for both customers and agents.

The disconnect between what customers expect and what traditional contact centers deliver is becoming increasingly untenable. According to Capgemini’s World Retail Banking Report 2025, only 24% of customers are satisfied with their bank’s contact center interactions. Customers cite long wait times (61%), inconsistent communication across channels (65%), and gaps in real-time updates between digital and in-person interactions (63%) as major sources of frustration​. These pain points are equally prevalent in the insurance sector, where policyholders frequently encounter delays when filing claims, updating policy details, or seeking assistance during critical life events.

Operationally, these challenges extend beyond customer experience. Many financial services firms continue to operate contact centers where over 80% of an agent’s workday is consumed by repetitive, manual tasks, leaving little room for value-driven customer engagements​. For insurers, this often means agents spend excessive time manually verifying policyholder information, processing claims, or handling routine inquiries—tasks that could be streamlined through automation. Similarly, in banking, less than 10% of agent time is spent on revenue-generating activities such as up-selling and cross-selling, leading to missed opportunities for business growth​.

The demand for change has never been more pressing. Financial institutions must move beyond incremental improvements and embrace AI and generative AI (GenAI)-powered contact centers that blend automation, real-time analytics, and live agent support. By integrating these capabilities, banks and insurers can significantly reduce operational costs, improve customer satisfaction, enhance compliance monitoring, and strengthen fraud detection efforts. Those that adopt AI and GenAI-based solutions will be well-positioned to turn their contact centers from cost-heavy service departments into strategic hubs of customer engagement, operational efficiency, and revenue generation.

AI-powered contact centers: The key to efficiency and growth

By integrating advanced AI and GenAI technologies, banks and insurers can create smarter, more responsive, and cost-effective contact centers. From real-time insights that improve self-service interactions to automated workflows that streamline post-call processes, these tools are redefining efficiency and service quality.

Improving efficiency through real-time speech recognition

Efficiency is at the heart of a well-functioning contact center, yet traditional workflows burden agents with manual notetaking, post-call documentation, and slow information retrieval—all of which extend call durations and reduce productivity. AI-powered real-time speech recognition technology is improving agent workflows by providing instantaneous transcription, automated notetaking, and intelligent response suggestions, allowing agents to focus on engaging with customers rather than administrative tasks.

NVIDIA® Riva, a collection of GPU-accelerated multilingual speech and translation microservices, enables firms to create or fine-tune open-source automatic speech recognition (ASR) models to better comprehend sector, function, and firm nuances to generate highly accurate transcriptions at low cost. By leveraging NVIDIA® NIM and NIM Operator for scalability, financial institutions can seamlessly deploy instant transcription across large-scale contact centers without disrupting existing workflows.

Bolstering customer satisfaction through deep, real-time insights

Contact centers incorporating GenAI capabilities are revolutionizing self-service by delivering context-aware, human-like interactions that go past scripted responses. Unlike traditional solutions, contact centers powered by capable state-of-art large language models (LLMs), trained on enterprise data, can better understand customer intent, retain long-term context, and provide real-time, personalized support.

With AI21’s Jamba—a hybrid state-space and transformer LLM—GenAI can provide advanced sentiment analysis, low-latency response times (<500ms), and deep contextual understanding that adapts as the conversation evolves. NVIDIA® NeMo can be used to customize the models with domain knowledge. Once in production, model performance can be maintained with NVIDIA® NeMo microservices to curate new business data and user feedback, fine-tune and evaluate the model, connect with Retrieval-Augmented Generation (RAG) pipelines, and guardrail the model’s responses. Furthermore, NVIDIA® NIM can help scale latency and throughput, optimizing the delivery of GenAI-driven insights. 

GenAI-powered self-service also ensures seamless omnichannel experiences, enabling smooth transitions between chat, voice, and video interactions while maintaining context. By integrating intelligent automation and real-time insights, banks and insurers can provide faster, more relevant support—boosting efficiency while strengthening customer loyalty.

With real-time transcription and GenAI-driven insights, agents receive instant customer context and recommended responses, allowing them to resolve inquiries more efficiently. This technology also automates post-call work, generating summaries of key details and next steps—tasks that traditionally take several minutes per interaction. As a result, financial institutions can increase overall agent productivity by an average of 14% and by 34% for novice or lower-skilled workers, optimize call handling times, and empower agents to deliver faster, more personalized service.

Simplifying AI and GenAI adoption with a Contact Center-as-a-Service platform

A Contact Center-as-a-Service (CCaaS) platform streamlines the adoption of AI and GenAI capabilities by eliminating the need for complex infrastructure. With plug-and-play integration, firms can rapidly deploy AI-driven real-time speech recognition and GenAI-powered agent assistance without disrupting existing workflows.

Zuqo’s CCaaS platform can be used to accelerate deployment by orchestrating automation, live-agent interactions, and post-call workflows in a seamless environment. This allows firms to enhance customer engagement and agent productivity without the need for major technology overhauls.

With built-in scalability, Zuqo’s platform supports multi-region and multi-language operations, enabling financial institutions to expand without technological bottlenecks. Its API-driven architecture ensures effortless integration with existing CRM, compliance, and fraud monitoring systems, allowing AI-powered enhancements to fit naturally within current workflows.

Use case: Rapid fraud resolution in financial services

A customer calls their financial institution’s contact center after noticing an unauthorized charge on their credit card account. Instead of navigating frustrating hold times or being transferred multiple times, they are quickly connected to a live agent equipped with AI and GenAI-driven support tools.

As the customer explains the issue, real-time speech recognition transcribes the conversation, instantly analyzing intent and retrieving relevant account details. The GenAI-powered system assists the agent by surfacing next-best actions, allowing them to immediately credit the disputed amount while the fraud investigation takes place. With a single click, the agent efficiently cancels the compromised card and issues a new one, ensuring a swift resolution without requiring the customer to call back or complete additional steps.

Once the call ends, AI-driven quality assurance automatically reviews the interaction within 150 seconds, assessing over 60 compliance and service quality indicators to ensure a high standard of support.

The future of contact centers is finally here

The financial services industry is undergoing a profound shift, where traditional contact center models can no longer keep pace with rising customer expectations and increasing operational inefficiencies. AI and GenAI-powered solutions provide the opportunity to transform these challenges into competitive advantages, enabling banks and insurers to deliver faster, smarter, and more seamless customer experiences.

By integrating GenAI-driven self-service, real-time speech recognition, and automated workflows, financial institutions can enhance agent productivity, improve fraud resolution, and ensure regulatory compliance—while reducing costs. Contact Center-as-a-Service platforms make this transformation even more accessible, providing a scalable and easily integrated solution that eliminates the barriers to AI and GenAI adoption.

As the demand for efficiency, security, and personalized service continues to grow, financial institutions that embrace contact centers powered by AI and GenAI will position themselves as industry leaders. By modernizing their approach, banks and insurers can future-proof their organizations in an increasingly digital world.

The jump from cloud FinOps to on-demand consumption FinOps is underway, it’s time to embrace the evolution and unlock greater business value.

Cloud FinOps was built around utilizing the public cloud to manage financial investments and has become a standardized part of today’s technology landscape. However, FinOps cloud cost management is no longer enough – it’s time to evolve. To do so, it’s worth considering what cloud FinOps is currently defined as to better highlight our solution.

The FinOps Foundation, a project to advance the community who practices FinOps, defines cloud FinOps in their Framework as: “an operational framework and cultural practice which maximizes the business value of cloud, enables timely data-driven decision making, and creates financial accountability through collaboration between engineering, finance, and business teams.”¹

However, with the growth of on-demand technologies like SaaS and Gen AI, FinOps needs to go further. And with this evolution in business and technology comes a new focus on operating expenses (OpEx), particularly in how these technologies affect cost. Companies need to evaluate the value of OpEx spend, while FinOps need to underline the benefits.

To put it simply it’s time to ask: What’s the cost of a click?

It is undeniable that cloud FinOps has proved its worth both quantitatively and qualitatively in terms of cutting the fat out of technology investments. It’s driven elevated business value and has become a staple discipline across organizations.

Still, there are a number of limitations that contemporary FinOps practices have run into. These primarily circle around common themes such as isolation and lack of strategic input. And with companies seeking to manage their OpEx spend, ensuring that FinOps and on-demand technology integration continues can become complicated by a lack of visibility among FinOps teams.

Let’s examine these challenges in more detail. 

Isolated FinOps Teams 

There’s no denying FinOps teams do a great job in their areas of influence, however, those areas are often too limited. This is especially true when executive sponsorship wanes and other priorities take precedence. The focus on cloud FinOps seems to return only when a negative cost incident occurs, and this only further isolates these teams. 

Low demand management influence 

Even the most successful cloud FinOps capabilities are limited in their influence in demand management. Specifically, in how FinOps teams influence acquiring new technologies and resources to support the growth of the discipline. Sure, they can highlight cost implications and even help shape cost-effective solutions, however, they can rarely influence demand management directly – especially as part of a wider, end-to-end system.  

Limited strategic reach 

FinOps teams usually report to a technology leader who isn’t likely sitting in the C-Suite. Naturally, this limits the influence on strategic decision-making that FinOps has. This is a compounding problem, as other areas of the business group beyond the IT organization are driven by C-Suite priorities and bridging that gap is hard – if not impossible – in a large enterprise.  

Isolated from other initiatives 

To add to this gap, initiatives from senior leadership most often do not happen in concert with FinOps teams. The consequence can impact how architectural principles are re-designed, or result in frugal investments made in architecture that should support FinOps. It translates into a lack of FinOps integration with the wider business. The result? Unplanned or unexpected complexity later.  

Singularly focused on cloud and public cloud services 

In the current market, there is an ever-increasing variety of consumption-based technologies in use. FinOps teams who are only using public cloud are limiting the potential value they can generate. This in turn results in the web challenges we’ve described. It’s precisely the place where traditional cloud FinOps stumbles, and where exactly is that leading us?

While the FinOps Foundation is widening their scope to include SaaS and AI, within the overall FinOps community, these concepts are still in their infancy. This is true both in terms of implementation but also integrating them into current FinOps practices.

However, we believe this is a vital moment to start the shift towards acquiring new technologies and re-defining goals. This is reinforced by the fact that there is an increasing amount of tooling available to not only identify, but also to introduced automated optimization of various FinOps processes. And with the continued adoption of machine learning and AI (which is already rolling out as AI for FinOps) – this will only accelerate.  

Welcoming the new era of FinOps:

Cloud Consumption On-Demand 

From the very beginning, the rise of Cloud Consumption On-Demand in FinOps must consider all consumption-based technologies and business strategies. This includes everything from SaaS, Gen AI, AI infrastructure, and other cloud FinOps services. In turn, these technologies must be introduced to encompass the entire business value chain.

To do this, FinOps teams need to be challenging traditional capital expenditure (CapEx) governance systems. In a world where a growing number of technologies are purchased under OpEx scenarios, this can be complex. In other words, FinOps teams must not only prove long-term financial benefits but provide a solid short-term business case that tangibly shows how Cloud Consumption On-Demand technologies create cost savings and better efficiency.

It also means re-defining how FinOps is perceived. From their current periphery, FinOps teams must harness on-demand technologies to demonstrate their value with faster and automated solutions, reduced expenses, and elevated data-led decision making. This will help showcase the strategic value FinOps teams are creating.

Additionally, FinOps teams need to underline that the challenges they face are business problems. It is not a question of managing IT systems or technologies. This is a matter of delivering better business value thanks to the benefits of on-demand technologies, such as accurate budgeting and elevated forecasting results.

FinOps teams can also deepen collaboration across business units with on-demand technologies, For example, providing more flexible scalability that best fit the immediate business need. This can help keep various disciplines, like finance and engineering, connected and transparent in terms of financial accountability via integrated tools and platforms with accurate, real-time data sharing. 

Transforming company culture

While harnessing on-demand technologies and more modern FinOps practices are vital, it is education that will really make an impact. FinOps teams have the opportunity to transform a company’s culture by pushing for stronger education around on-demand technologies and their potential. Such a step is not only desirable but business critical. 

These cultural practices that on-demand FinOps can create will help spark greater levels of integration across business lines. This translates into the reduction of silos across the organization. Ideally, the new FinOps will fully integrate with demand architecture, purchasing, on-demand technology, reporting, analysis, finance forecasting, and more.

It’s time to evolve the FinOps framework 

This new world of FinOps will play a far more active role. Powered by on-demand technologies, the discipline can escape the limitations that reliance on public cloud has introduced and in turn, generate greater value. Most importantly, it gives FinOps teams deserved recognition as fundamental players in modern business practices. 

Do you know the cost of a click?

With Cloud Consumption On-Demand, we can help diagnose the potential of FinOps has in store for companies, while supporting them with action plans to help them make it a reality.

It’s time to welcome in the new era of cloud FinOps…

… It’s time to know the cost of a click.

Reference: 1. FinOps Foundation Framework

The future of public administration lies in partnerships—not silos—with citizens, businesses, and civil society. In an era of rapid digital transformation, while the guiding principle of providing accessible, inclusive and high-quality public services remains fundamentally unchanged, the way public administrations are creating value for their citizens is undergoing a profound evolution.

As technology evolves and societal challenges grow more complex and interconnected, traditional siloed structures are increasingly being replaced by dynamic ecosystems where value co-creation is critical to the success or failure of public interventions.

In 2021, 85% of public administrations in Europe were already using some form of co-creation to innovate public-service delivery. Today, this approach has become a widespread foundational principle. Key technological enablers are driving this shift, empowering public administrations to move towards a collaborative approach of public service delivery that brings together governments, businesses and citizens to address challenges more effectively. From leveraging interoperability to dissolve boundaries and advance data-sharing ecosystems to the rise of GovTech, proactive service delivery and the transformative potential of government AI, these key trends are laying the groundwork for a smarter, more inclusive and efficient public governance designed to meet the demands of modern, interconnected societies.

Five trends in public administration

Time for action in an increasingly interconnected world

To fully harness the potential of these digital trends, public administration leaders must adopt an action-oriented approach. A combination of political commitment to digital transformation, inter-agency collaboration and leveraging robust PPPs to bridge resource gaps and accelerate innovation will be key. Together they will help to overcome budget constraints, siloed institutional frameworks, cultural resistance to change and complexities in measuring and reporting progress that still afflict public administrations worldwide. While strategically investing in cutting-edge technologies like AI, leaders must also champion a culture of continuous learning and upskilling among civil servants, ensuring they are equipped to leverage effectively these emerging tools. Ultimately, aligning digital strategies with citizens’ needs through human-centered service delivery will enable administrations to build trust, improve efficiency, and deliver meaningful public value in an increasingly interconnected world.

Forget chatbots. The age of the agent is here. Imagine a digital workforce that understands, empathizes, and anticipates customer needs as a trusted advisor – a network of AI agents collaborating to deliver truly human-centric experiences.

This isn’t science fiction; it’s the dawn of the Agentic AI era, and it’s poised to revolutionize customer interactions. Market.us is projecting the global Agentic AI market will be valued at $196.6 billion by 2034, a dramatic leap from $5.2 billion last year. This exponential growth is not just exciting; it signals a fundamental shift. While the possibilities are vast, companies must move beyond simply creating “cool agents” to building robust, collaborative systems.

Agentic AI is rapidly evolving, and the conversation needs to shift towards building networks of interconnected AI agents. This next stage, focusing on multiagent systems, is where real value will be unlocked. 

Next-level hyper-personalization: The game changer 

The true power of multiagent systems lies in their ability to deliver hyper-personalized experiences. Imagine AI agents seamlessly orchestrating across different business areas, instantly accessing client information to tailor interactions in real-time. This level of hyper-personalization, incorporating individual preferences, creates a genuine sense of personal connection. 

Multiagent systems represent the next evolution in personalized interactions. We’ve moved beyond deterministic chatbots and automated processes to a realm where embedded generative AI enables faster, more personalized interactions that build loyalty and connection. The impact is already evident: according to the Capgemini Research Institute, 31 percent of organizations using generative AI see faster response times, and 58 percent anticipate further improvements. 

Efficiency and beyond: Connecting agents across departments 

Beyond enhancing customer experience, connecting agents across departments drives efficiency and productivity through automated, complex workflows. The ability for agents to communicate and operate seamlessly at faster speeds across departments unlocks significant potential. 

This also expands service capabilities. For example, overcoming language barriers in global call centers becomes possible with multilingual digital agents. Research indicates that 60 percent of consumers would pay more for premium customer service, highlighting the value of these enhanced capabilities. Google’s Customer Engagement Suite (CES) provides the AI technology and natural language processing (NLP) that can provide enhanced customer experiences.  

Connecting agents and data: Unlocking deeper insights 

Multiagent systems generate valuable data on information and conversations, which, when shared, provides a deeper understanding of customer behavior and trends. 

This data spans various departments – sales, order management, supply chain, ERP, and marketing – highlighting that inquiries rarely fit neatly into departmental silos. Agents need to be able to access data across these silos is crucial for providing cohesive responses to complex customer questions. 

This is why cross-department collaboration is crucial. Agents need seamless handoffs and access to different departments so that when a person engages with them, the conversation continues without waiting for the next agent to be updated. 

However, simply opening up data is not enough. Robust security protocols are necessary to ensure that not all information is accessible to every agent. Agents must pull information in a way that maintains visibility, requiring a deep understanding of systems for effective deployment. Data security and privacy are paramount. Accessing various data sources requires clear guidelines and governance to ensure compliance with existing data rules. 

Agentic change management: Blending the human workforce with the “digital workforce” 

Ideally, digital and human workforces will seamlessly blend, working in unison on daily tasks and customer interactions. Generative AI will continuously learn from feedback and algorithms, while large language models adapt. However, potential biases must be addressed to ensure fairness. 

Companies must also address the impact of multiagent systems on the human workforce. Clear communication early in the process can prevent resentment toward AI agents. Reassuring employees is a crucial part of change management. If employees fear job losses, they will be less inclined to engage with companies using AI agents. Multiagent systems offer exciting possibilities, but everyone must be part of the solution to maximize the benefits. 

Building a resilient agentic infrastructure 

Agentic AI does not mean creating a single, all-encompassing agent. Companies must prioritize resilience. Humans have bad days, and so can AI agents. If a single agent fails, the entire operation can grind to a halt. A multiagent system allows agents to focus on specific areas, ensuring that if one fails, others remain unaffected. 

The challenge for companies lies in the complexity of the infrastructure required for seamless agent communication. While technology is increasingly sophisticated, the talent to make it work is scarce. Companies need the right skills to build and effectively operate these agentic systems. 

Google’s Agentspace is an orchestration platform that allows companies to deploy agents easily. The Google ecosystem integrates seamlessly with any system, ensuring smooth information flow, regardless of whether a company is using Google applications and infrastructure. 

Working with Google Cloud, Capgemini can support customer service transformation that creates seamless, quality interactions that deliver an exceptional level of service, support, and delight to all stakeholders. Advanced AI capabilities and scalable infrastructure means Google Cloud can build and deploy intelligent virtual agents, enhance agent productivity, and personalize customer experiences easily. We can leverage the power of Google’s Customer Experience Suite to innovate for growth and reinvent business models to unleash what is possible. 

Join us at Google Cloud Next to discover how we’re helping companies embrace the agentic era and benefit from the intersection of innovation and intelligence.

Call centers are seeing gains but reliability and consistency need to be a focus. Adopting a copilot approach is the best way to ensure real efficiencies and positive customer experience. 

It has been said that AI agents could be a multi-trillion dollar opportunity. Intelligent software agents capable of learning to manage actions and tasks have the potential to transform almost everything. Work and life will be impacted by the drive for productivity and efficiency. But AI agents will also democratize access and help overcome barriers to empower more people and drive innovation.

The road to agentic AI is still being built and there will be many routes to explore but, today, one of the biggest pushes is coming from the telecommunications industry. Call centers have been early adopters for this kind of generative AI because it is a natural evolution from bots. Existing chat features were interesting but they do not always work well, or customers were so annoyed by the menu systems on phones that they were unhappy before they even spoke to someone.

Moving beyond bots

Generative AI brings a much better experience to the call center structure, and it enhances existing technology. For example, Google Customer Experience Suite (CES) was built on its Contact Center AI and enhanced with generative AI technology. It has better engagement with customers in both the chat channel and live. With the emerging capabilities of large language models (LLMs) and the growth of companies like OpenAI, AI agents can take on expanded tasks.

Creating a multi-modal experience allows an AI agent like Google Gemini to intake text, visuals, and audio, and add a communications layer through actual voice and text-to-voice features that are extremely realistic.

Combining Gen AI, language features, the ability to understand a vast amount of context instantly, and better and more human communication with text-to-voice capabilities creates systems with huge potential.

Enhancing the AI agent

We have recently launched the concept of thinking models that are capable of handling much more complex tasks. This is achieved through reinforcement learning based on human feedback, which means these models can actually think, process, and approach problems from multiple angles and explore different paths to find the best solution. It is very reminiscent of how a human would work to solve a problem.

Agentic AI has the capability to not only understand what a customer needs but to communicate in our own language with the right nuances and even slang. Communication is there. The thought process is there. The ability of AI agents to think through problems at length is there. And they bring the ability to use tools during interactions.

For example, a customer calls in with multiple inquiries. The AI agent can quickly understand the intent of the call so there is no longer a need to sift through menus or listen to a bunch of options. Because the intent is read in the early stages of the call, the problem resolution process operates better, as the AI agent has the information to solve the problem and the tools to execute it.

The right AI team

After detecting the true intent of a customer call, a master AI agent can act as the interactive layer with a customer, while simultaneously accessing a team of subagents to delegate tasks. The subagents can specialize in different areas, like billing issues or new installations. There is no more waiting on hold to be transferred to a different department or a manager. The master agent can access a whole host of tools and know what it needs to take action.

For example, a customer may want to process a payment. The master agent can identify the request and decide how to proceed. It can give a credit, research a billing discrepancy, or initiate other searches to complete the request. It can call different APIs to get information, update the account, and process the bill. With access to tools, there is really no end to what an AI agent can do.

These reasoning capabilities and tools mean agents can do very similar things to humans. However, it is still early days in the process and there are concerns to be addressed. Reliability and consistency are two factors. The monitoring and evaluating are improving to help ensure the responses and decisions by AI agents are correct.

Improving the call center experience

We worked with one telco client to deliver better knowledge searching, to leverage LLMs to use new methods of data acquisition summarization. The goal was to make technical documentation more accessible so when someone calls to troubleshoot a modem, for example, the answer is readily available.

Call centers are also a common sales channel. Agents can provide additional information or offer specific deals. That requires the agent to understand the needs of the customer, align them with a product, make an offer, address objections, and close the deal. Now an AI sales agent can interview the customer to understand the needs and wants and match them with potential solutions. They can even address objections and concerns to help get to the sale.

Copilots: Finding the agentic balance

According to a recent Capgemini Research Institute report, being an agent is not an overly satisfying career choice, with only 16 percent of human agents surveyed report overall satisfaction with their roles. They face a number of pressures, from rising customer expectations to inefficient systems and a high attrition rate. There are efficiency gains to be made by employing AI agents that can help humans do a better job. In addition, AI agents can help resolve issues more quickly so the customer and employee have positive experiences.

This is why the copilot effect is a popular AI agent option. Google has Agent Assist to support live agents to resolve queries and issues more quickly. It is like having an expert in the room at all times with a call center agent. For example, the human agent can use it to help digest what is being said, with information automatically appearing on dashboards to assist with the call resolution. The copilot can also provide real-time assessments of the sentiment of the caller. Now the human agent has prompts with potential resolutions, rather than having to bounce between different systems for information or consult with a manager.

The human in the middle

So the concept of the human in the middle is very important. AI agents are a powerful tool meant to enhance experience, but sometimes a model can hallucinate or produce an error – and a company is responsible for an AI agent’s output. That means companies have to own the net result. So employing copilots with the human in the middle is happening even in new call centers. Once a system is proven, the role of AI agents can expand but, since call centers have a major impact on customer experience, there needs to be a high level of comfort with the system.

Call centers that use Google Customer Experience Suite (CES) engage customers with generative AI for many tasks, like determining what a client needs and other lower-level processes, to make calls more efficient and get to resolution quicker. AI agents can, for example, engage with back-office operations so humans can focus on more high-value tasks.

It takes time for companies to be comfortable with exploring generative AI solutions.  Companies need to focus on the business case and ensure innovation results in efficiency and savings.

Working with Google Cloud, Capgemini can help companies move into the agentic future. We can help companies build a competitive edge with agents to drive real customer service transformation. Google Cloud’s advanced AI capabilities enable businesses to build and deploy intelligent virtual agents easily. It is time to create, frictionless environment to scale agents where everything supports the needs of the organization and its customers.

Join us at Google Cloud Next to discover how we’re helping companies embrace the agentic era and benefit from the intersection of innovation and intelligence.