According to James Saunders, co-founder and CTO at Anti-Money Laundering (AML) platform RelyComply, developers and web operations (WebOps) specialists face a challenging year as organisations strive to drive more value from their ICT investments, ramp up cloud migration, fortify cybersecurity, and address sustainability and environmental imperatives. Here are five software skills we believe will be essential in 2025.
- Cloud development and operations
From remote work to digital transformation, the cloud underpins almost every business’s IT strategy today. It allows businesses to rapidly develop and deploy market-ready systems while taking advantage of the inherent flexibility, scalability, and resilience of as-a-service computing models.
Yet almost 100% of organisations say implementing cloud technology is tricky, with managing costs and wastage being among the most significant challenges. Digital experts who can help enterprises use and provision cloud-based services efficiently will be in high demand.
These experts will know how to leverage Infrastructure-as-Code (IaC) tools to automate development and DevOps. Engineers who understand the challenges of hybrid cloud environments (with public and private clouds and on-premises resources in use) will also be needed more than ever.
- Cloud Financial Operations
Choosing software is a business decision that can make or break an IT department, especially in this time of stretched tech budgets. Cloud Financial Operations (FinOps) experts help companies ensure their cloud investments drive revenues rather than rack up unnecessary expenditures.
Financial optimisation skills that will be important in 2025 include:
- Addressing additional services beyond the core infrastructure that can incur further costs.
- Identifying hidden fees – including migration costs, extra support deals, termination fees or region-specific pricing.
- Comparing benefits of cloud solutions against on-premises centres for data storage.
- Using only data budgeted for and increasing or decreasing storage needs as appropriate.
- Minimising instances that are not delivering value about their storage space or those that are idle.
- Making the most of cloud providers’ commit discounts and avoiding over- or under-provisioning.
- Automated infrastructure management
Google developed Kubernetes, a portable open-source solution for running, deploying, and scaling containerised applications or clusters (containers across multiple servers) and ensuring consistency across environments. It enables companies to eliminate many manual DevOps processes.
Developers are becoming more familiar with honing this architecture’s automation. This involves designing and implementing clusters, managing communications between containers and services within them (such as network policies or load balancing), and monitoring cluster health through access controls, storage volume metrics, and more.
- Strengthening fortifications with DevSecOps
Cybercrime threats like malware, ransomware and denial-of-service attacks are evolving rapidly, and new vulnerabilities are constantly emerging. New dangers lurking around the corner include connected internet-of-things (IoT) devices, complex cybersecurity standards, the effects of blockchain, and identity fraud through biometrics.
DevSecOps is becoming more widespread, putting security in the hands of developers. Developers must understand how to configure systems according to certifications such as ISO 27001. They must also master cybersecurity skills, including remote access protocols, VPN usage, firewalls, intrusion detection, and penetration testing.
- Facing an AI future
Artificial intelligence (AI) remains a space where experienced developers are in strong demand. Not only does AI play a key role in tasks such as uncovering patterns in big data sets in applications of AML monitoring, but it is also becoming a powerful tool to increase automation and complement developer productivity.
AI-based agents are beginning to be able to automate tasks that previously human operators would have to do without long training periods or complex setups found in the previous generation of RPA (Robotic Process Automation) tools. This will allow human operators to focus on more valuable tasks, eliminating the mundane and allowing them to be more efficient and effective.
Code generation via generative AI also offers two distinctly interesting possibilities. Firstly, it can be used to allow non-technical (or less-technical users) to perform queries on data and similar systems without having to learn skills like SQL or Python. This is done by letting the users type in natural language what they want and generating the underlying query.
There has been a wealth of new coding assistant tools for developers, and while they have certainly shown value in some cases, especially around shorter coding tasks, it is far from certain that these will generalise more complex software systems. It seems certain they will be a critical part of the modern development landscape, but the gap between the expectations set and the daily reality of these tools is still large.
