A Changing Task Environment
A study from the University of Zurich reports that developers use an average of 16 tools in the course of their day, and spend their time switching between a wide variety of tasks. The resulting complexity can be exhausting for individuals and can result in inefficiency for the organization. Amanda Silver, general manager for Microsoft’s platform engineering team, believes that platform engineering can improve the developer experience and optimize the software development lifecycle.
Platform engineering is already becoming crucial to IT development. According to one report, more than 80 percent of developers spend less than 30 percent of their time writing code–and sometimes as little as 12.5 percent. It’s estimated that by 2026, 80 percent of software engineering companies could be using platform teams as internal contributors for tools for application technology, reusable services, components, and more.
Experts predict several trends that will affect the development of platform engineering.
A Shift Toward Serverless Architecture
Serverless architecture describes two different but overlapping areas: Backend as a Service (BaaS) and Functions as a Service (FaaS).
Backend as a Service applications incorporate third-party cloud-hosted services and applications to manage server-side logic and state. These applications may use, for example, cloud-accessible databases and authentication services. BaaS services can help to streamline and automate backend development tasks and make it easier for developers to focus on creating applications.
Functions as a Service includes applications in which server-side logic is written by the application developer but run in ephemeral, event-triggered stateless compute containers and managed by a third party.
These two types of serverless are often used together, for example in Google’s Firebase BaaS database, which has FaaS support through Google Cloud Functions for Firebase.
Benefits of serverless architecture include enhanced efficiency and scalability, reduced operational overheads, and quicker deployment cycles. At the same time, challenges remain with regard to vendor lock-in and the optimization of performance.
Integration of Security with DevOps
DevOps and security teams often have different cultures and different ways of working. DevOps can sometimes see security as a barrier to speed and agility, and therefore productivity.
But both DevOps groups and security teams share the goal of protecting their cloud infrastructure from threats. Platform engineering can help to bridge the gap. Automation can facilitate seamless embedding of security measures throughout the development lifecycle, and can help to maximize both performance and security.
Minimizing DevOps security tasks through customizable policy templates and just-in-time access that doesn’t disrupt workflow can help to make the jobs of both DevOps and security teams easier and more efficient.
Edge Computing
Edge computing is a type of distributed computing in which data collection, processing, and storage take place near the source of that data – that is, at the edge of the network. By moving computation away from data centers, edge computing can help to minimize latency, as only the most important data is transferred to the central data center for processing. This can enable real time processing and improve efficiency, especially for data intensive applications.
With edge computing, connected products, mobile phones, and network gateways can provide services such as content caching, service delivery, IoT management, and persistent data storage, as well as perform tasks on behalf of the cloud. This can result in better response times and transfer rates.
The research firm Gartner predicts that by 2025, the percentage of enterprise-generated data created and processed outside of a centralized data center will increase from today’s 10 percent to 75 percent.
Quantum Computing
Quantum computing uses specialized hardware to leverage quantum mechanical superposition and entanglement, in order to perform certain calculations exponentially faster than ordinary computers.
Although the field is largely experimental at present, quantum computing shows promise in areas such as encryption-breaking and more. Both public and private entities are investing in quantum technology. Indeed, in 2022, the Association for Computing Machinery (ACM) reported that the European Union was spending €1 billion on quantum technologies. At this point, one major goal is bring the technology out of the realm of theory and into practical usage.
Integration of AI, GenAI, LLMs, and Machine Learning
Artificial intelligence shows unrivalled potential for improving speed and efficiency through automation. Some tasks to which AI is particularly well suited include real-time data analysis, which can help platforms to dynamically adapt to user behaviors, code change management, security management, and software testing. Building AI into platforms can enable companies to optimize resource allocation, personalize user experiences, and automate complex tasks like never before.
AI’s data crunching capabilities also have the potential to revolutionize monitoring and performance during operation. A smart monitoring system can gather and analyze real-time performance data, making it possible to identify resource-heavy sections and find and fix problems quickly and easily.
Generative AI (GenAI) can improve user experience through natural language. It also has the potential to democratize worker access through prompt engineering using natural language. This can improve not only access, but also quality of output. Experiments have shown that using large language models for tasks such as coding and marketing can result in performance improvements of between 30 and 80 percent.
According to Gartner, in 2023, less than five percent of enterprises used generative artificial intelligence APIs or models. However, Gartner predicts that by 2026, that number will have risen to more than 80 percent. Industries in which demand for GenAI is increasing include life sciences, legal and financial services, healthcare, and the public sector.
Combined with machine learning, AI’s potential seems practically limitless.
At the same time, it will be important to recognize and plan for the mitigation of the risks associated with AI, including breaches, inaccurate or unintended outcomes, process errors, and even interference from malicious actors. Gartner proposes a system of trust, risk, and security policies (AI TRisM) aimed at ensuring the trustworthiness, reliability, robustness, efficacy, and security of AI models. It’s in organizations’ best interest to adopt a transparent system of AI TRiSM. Gartner predicts that by 2026, companies that do will see their AI models achieve a 50% improvement in terms of adoption, business goals and user acceptance.