Pragmatic analysis of 2025 industry trends by ELEKS' technology experts
AI/GenAI will continue to find its place in business operations; data strategy and proper data governance are prerequisites to success in AI/GenAI implementation, especially in AI agents development. Cloud computing is still a big enabler for either innovations (e.g. GenAI) or optimisations for many businesses.
Security concerns need to be considered when you work with AI, cloud, and data. Regarding software developers' productivity, many companies are focused on developer experience (e.g., developer platforms, AI-augmented development, streamlined processes), which is a multifaceted approach to ensure that engineers spend time on focused work and value creation activities.
Artificial intelligence
Artificial intelligence is everywhere, it's becoming as basic as electricity or the internet. Gartner and Deloitte both agree that it's only going to get bigger in 2025, especially generative AI. This technology is moving from a tool for automation to becoming the cornerstone of business strategy and innovation. This transition manifests in several ways.
Generative AI’s growing impact
ELEKS' expert evaluation:
The most significant shift in AI deployment comes from maturing of generative AI. This subset of AI will become deeply embedded in many enterprise operations. It's not just about introducing new systems; it's also about addressing the legacy systems many companies still rely on, which also cost them a lot of money to keep running. The good news is that GenAI can help make these old systems work better.
GenAI increasingly becomes engaged in everyday work situations, from financial analysis to deciphering medical images and more. But perhaps most importantly, GenAI changes the way we learn and use information in the workplace. Instead of spending hours searching through documents or asking colleagues for help, workers can quickly find answers to their questions using GenAI.
Forget about repetitive and time-consuming tasks and leave them to GenAI so that employees can focus on tasks requiring creative approaches and critical thinking.
Companies can effectively optimise costs by using GenAI for tasks like automation, legacy systems’ modernisation, resource management, and more.
The AI-driven personalisation and round-the-clock customer support boost customer satisfaction and retention.
Testing new concepts and creating prototypes can be accelerated with AI, reducing time-to-market for new products and features.
The rise of AI agents
ELEKS' expert evaluation:
Agentic AI should be the bullseye of our focus in 2025. Multi-agent workflows in particular–which are highly complex and require governance–should present a rich opportunity for us in 2025. Creating automation in a front-office or back-office function will require specialised agents working together seamlessly to streamline or completely automate complex business workflows.
Unlike chatbots or copilots, agent AI acts independently, instead of responding to prompts, it chooses which actions to perform according to the goal set by a person.
Agentic AI systems can work collaboratively and make independent decisions while allowing for human oversight and refinement of their decision-making processes.
Superagents will take care of more effective interactions of AI systems.
AI agents combined with spatial computing will create systems that naturally understand and respond to our gestures, speech, and contextual cues, eliminating the need for specialised commands.
AI agents can enhance the performance of LLMs, enabling their use on small devices (edge AI) and offline.
Shift to small language models
ELEKS' expert evaluation:
We are observing a trend toward smaller, specialised AI models that are designed to meet specific business needs. It's no longer just about large AI projects; instead, the focus will be on finding practical ways to integrate AI into existing company operations.
LLMs use many resources and aren't always necessary for every task. Small language models (SLMs) can be more efficient and cost-effective for simple tasks. For example, summarising an inspection report doesn't need a large model; a smaller model trained on similar documents will work just fine.
LLMs are designed to answer questions but not for doing particular tasks. SLMs can be trained on selected data to solve specific problems.
Smaller models can be run on-device, enhancing data security and reducing reliance on the cloud.
SLM rises as the greener alternative with lower computing needs, smaller energy consumption, and reduced carbon footprint.
Nowadays, there are many different directions in AI, and most of them, due to various non-scientific or business sources, are considered the top priorities for 2025. The same was done in previous years with ethical AI, robustness in decision-making, precision in a generation, and sustainability issues indicated as key focus areas.
However, one challenge stands out, appearing in around 80% of cases related to modern GenAI technologies, especially the LLM-based agents. This type of problem lies in the intersection of AI and security and is about the i/o (input-output) validation or verification of the LLM-based agents. When talking about output validation, most applied specialists may initially mention model performance, such as different types of precision or LLM evaluation.
In my view, i/o verification is needed to avoid inappropriate agent usage (e.g., front-office chatbots), data leakage (e.g., prompt injections and indirect prompt injections), jailbreaks, and irrelevant data in the outputs. Currently, it is a challenge that leads to numerous problems, such as users bypassing system instructions or retrieving hidden data through indirect prompt injections (especially in retrieval-augmented generation systems). Unfortunately, there is no single, effective solution to comprehensively address these issues.
Cloud computing
The cloud computing landscape is transforming rapidly, bringing both opportunities and challenges for businesses worldwide. Let's explore the key trends shaping this evolution and what they mean for enterprises.
FinOps optimising cloud costs
ELEKS' expert evaluation:
What's particularly noteworthy for 2025 is the shift from reactive cost management to proactive value optimisation. When properly implemented, FinOps provides C-suite leaders with real-time visibility into the return on cloud investments–something that has been notably absent in traditional cloud management approaches.
My message to fellow business leaders is simple: if you haven't prioritised FinOps in your 2025 technology strategy, you're putting your company at a disadvantage. It isn't only about cost savings; it's about building a foundation for sustainable cloud operations and competitive advantage in an increasingly cloud-dependent world.
However, many organisations find it difficult to keep track of and understand their cloud costs. Here's where FinOps enters the picture. This collection of tools and strategies helps organisations get a handle on their cloud spending.
With FinOps, companies can ensure that every investment in cloud services delivers maximum value.
The framework provides insights into spending across different services, regions, and accounts.
FinOps makes managing the rules and requirements for cloud spending easier, including creating cost management policies, setting controls, and preparing reports.
FinOps effectively aligns cloud infrastructure with business objectives. Enabling strategic planning and agile implementation helps companies to adapt to changes in the market while also working towards their long-term growth.
It allows organisations to maintain sensitive data and critical workloads on private infrastructure while leveraging public cloud capabilities for less sensitive operations.
Businesses can choose where and how their applications run while complying with internal rules and industry regulations.
In an era of increasing data regulation, hybrid clouds provide the flexibility to keep sensitive data within specific geographic boundaries or organisational premises.
Hybrid clouds allow companies to allocate resources between private and public infrastructure, scaling up or down based on demand while optimising costs and performance.
No more waiting for information to travel to distant servers and back. When you need to make quick decisions, every millisecond counts.
Whether it's factory sensors or retail analytics cameras, these devices will work smarter and faster than ever before.
When data stays local, businesses don't have to worry as much about where their information travels.
Edge computing reduces energy consumption, supports more sustainable logistics, and helps minimise waste.
The best part? Edge devices are getting smarter. They're coming with built-in AI capabilities (called NPUs), which means they can make decisions on their own without always checking in with the cloud.
Think of it like having mini-computers everywhere, each handling its own tasks while still staying connected to the bigger network when needed. This hybrid approach gives businesses the best of both worlds–speed and cloud power.
Key strategic areas shaped by technology in 2025
Cybersecurity
Nowadays, cybersecurity isn’t just about protecting data–it is about staying competitive.
AI is changing the way attacks happen. In 2025, we will only see more AI-powered attacks, especially in phishing and social engineering. These attacks are becoming more frequent and are also becoming more convincing and harder to spot. To fight back, we need to do more than just defend ourselves; we should also use AI.
Cloud security will require a particular attention, as many enterprises operate across five or more cloud providers. The data shows that most cloud-related incidents stem from three primary sources: misconfigurations, inadequate monitoring, and weak access controls. Addressing these vulnerabilities will be critical to ensuring cloud resilience.
New regulations like DORA, the Cyber Resilience Act, and NIS2 are changing compliance requirements. Companies that must comply with these regulations will be running the projects and increase spendings to revise and improve their current processes. That would increase demand for cybersecurity services and experts in EU during 2025.
Companies are already implementing Zero Trust Network Access (ZTNA). Security operations are evolving toward a semi-autonomous model, where AI assists in alerts and threat prioritisation. Cloud-native security information and event management (SIEM) solutions have become standard, addressing the unique challenges of cloud environments. Organisations are also preparing for the quantum computing era by implementing post-quantum cryptography to protect against future threats.
Success in this new security landscape requires a comprehensive approach that goes beyond traditional defensive measures. Forward-thinking companies are:
- Developing robust AI impact assessments
- Building crypto-agile infrastructures
- Fostering collaboration between security, technology, and legal teams
Sustainability
In 2025, sustainability will be a key focus across various sectors, fuelled by the urgent need to address climate change, regulatory pressures, and the growing energy demands of emerging technologies like AI. Here's our overview of what sustainability is expected to look like this year.
Embracing carbon-free energy and sustainable operations
The intersection of energy transformation and business sustainability in 2025 is remarkable. As renewable power sources are gaining traction, and nuclear energy addresses the growing demands of AI infrastructure, we're seeing a fundamental shift in how businesses approach their energy strategies. The emergence of GreenOps exemplifies this transformation–companies aren't just tracking operational costs, they're integrating energy efficiency and carbon metrics into their core business decisions.
This isn't simply about compliance; it's about competitive advantage in a market where sustainability drives innovation and long-term value creation.
Tech-driven changes in the industry
AI in banking
The integration of AI into banking has moved beyond automation. Financial institutions use AI for sophisticated operations, from fraud detection to personalised financial advice.
Tech-driven changes in the industry
Further integration of AI
Just like in other fields, AI is shaking things up in healthcare, changing the way how medical professionals diagnose and treat patients. Perhaps most exciting is the rise of personalised medicine, where AI helps tailor treatment plans to individual patients, dramatically improving care effectiveness.
Hybrid care model is here to stay
The blend of traditional in-person medical visits and digital healthcare solutions will continue to create a more flexible and patient-centred experience. For many patients, virtual healthcare has already become the preferred option, especially for those seeking mental health services, as it allows them to receive care from the comfort of their own homes.
Tech-driven changes in the industry
AI in the public sector
AI has become a cornerstone of government modernisation, with a surge in adoption across public institutions. What makes this transformation particularly noteworthy is how governments are approaching AI implementation. Rather than wholesale adoption, as IDC reports, 60% of governments are embedding sovereignty requirements in their AI procurement, ensuring data compliance and operational resilience while mitigating risks.
Embracing digital citizenship
The era of lengthy bureaucratic processes is fading away as governments embrace technology to facilitate more intuitive and user-friendly interactions. By implementing low-code and no-code platforms, agencies can quickly develop and deploy digital services that better meet the needs of citizens.
Tech-driven changes in the industry
Changing energy sources
Renewable energy sources like wind and solar are about to surpass traditional coal power. At the same time, nuclear energy is making a comeback. Small Modular Reactors (SMRs) and Advanced Modular Reactors are also emerging as crucial solutions for providing clean, reliable power to meet this growing demand. These complicated energy systems need some pretty advanced software to keep them running smoothly, optimise their performance, and hook them up with the current power grids.
Focus on sustainability
The mining sector is embracing digital innovation while focusing on sustainability. Companies are implementing AI and machine learning not just to optimise operations, but to measure and achieve their decarbonisation goals. This digital revolution extends to supply chain management, where new AI-assisted systems are helping companies navigate increasingly complex market conditions.
Supply chain challenges
Both energy and mining sectors are grappling with supply chain challenges. From transformer shortages in the energy sector to ethical sourcing concerns in mining, companies are building more resilient supply networks. This includes diversifying supplier relationships, implementing robust performance management systems, and prioritising ethical sourcing practices.
Tech-driven changes in the industry
Monetising AI
Retailers will move beyond experimentation to full monetisation of AI capabilities. What makes this particularly interesting is the shift toward modular AI models specifically designed for retail applications. GenAI is revolutionising product content creation and delivery, while AI-driven loyalty apps are delivering increasingly sophisticated, contextualised offers to customers.
Physical stores become smarter
Physical retail is being reimagined through technology as well. Edge computing with onboard AI chips is reducing latency and improving store operations. Perhaps most notably, retailers are investing heavily in sophisticated loss prevention technologies that integrate data from multiple sources - RFID, cameras, traffic patterns, and inventory systems–to combat in-store theft more effectively.
Transform industry predictions into actionable implementation strategy
Understanding which technology trends hold genuine potential for your business requires more than market research. ELEKS' technical feasibility study service helps organisations transform promising innovations into validated business solutions with minimal risk.
Drawing on over 30 years of software industry experience, our certified experts help you evaluate emerging technologies against your specific business context.
Our team of R&D experts carefully assess technical viability, implementation requirements, and potential business impact before you commit to significant investments. We help you reduce technological uncertainty and build stakeholder confidence in your digital transformation initiatives.
Beyond technical assessment, we work with you to develop a strategic vision by mapping emerging technologies to your business roadmap.
Our comprehensive approach includes:
- Evaluating technical feasibility and business potential
- Developing proof-of-concept solutions to validate assumptions
- Creating detailed implementation roadmaps
- Seamlessly transitioning to full-scale development
As both advisors and implementers, we ensure that your investment in emerging technologies translates into real business value. Our end-to-end capabilities mean we don't just help you identify promising trends–we help you turn them into competitive advantages.
Related Insights
