GenAI Cloud Focused

Artificial intelligence has been a central topic among various public cloud providers in recent years. In 2024, this trend, particularly Generative AI, is further encouraged, reflecting companies’ commitment to this technology.

Different public cloud providers are rapidly expanding their toolsets for creating Large Language Models (LLM), accessing pretrained models, and offering managed services to facilitate and accelerate adoption. Examples include ChatGPT in Azure, Amazon SageMaker HyperPod, AWS Titan, Model Catalog in Azure AI, or Azure Prompt Flow, enabling companies to develop specialized models quickly without the need to access massive amounts of required data.

Additionally, to support this growth, we are witnessing the creation of a new generation of chips designed specifically for artificial intelligence, as well as optimized machine clusters for distributed training. Also, the emergence of vectorized databases and the expansion of the most popular databases to include this functionality.

Another significant aspect will be the extensive offering of virtual assistants like CodeWhisperer, Copilot, Q or Vertex AI Conversation, whose integration will extend across different work environments, from those offered by cloud services to those created and enhanced by various companies for internal use, all the way to integration with the most widely used IDEs in the market.

We can anticipate a democratization of generative AI models this year, driving the development of products that simplify the creation of applications based on this technology with more straightforward interfaces, such as “Low Code” or “No Code.”

The evolution of Generative Artificial Intelligence poses significant challenges in security. The ‘Generative AI Security Scoping Matrix‘ emerges as a comprehensive tool, categorizing GenAI solutions to identify security implications. Governance, legality, risk management, controls, and resilience become priorities in this context. In the case of third-party applications, attention is focused on governance and compliance, while for internal solutions, issues such as threats and resilience are addressed.

Adapting Identity and Access Management (IAM) for role-based access in GenAI becomes crucial, as well as designing resilient applications to meet availability requirements.

It seems necessary to have a guide to address specific security dimensions. In an environment where GenAI transforms the customer experience, these strategies offer a robust approach to ensuring security in the era of artificial intelligence.

Sustainability

Society is increasingly committed to the environment, and companies are no exception. Over the past few years, growing concern about climate change has led companies worldwide to develop different strategies to mitigate the environmental impact of their operations and promote more responsible practices in the management of technological resources. Major cloud service providers, such as AWS, Microsoft Azure, or GCP, are no exception, and they have committed to reducing their carbon footprint by signing the Climate Commitment. This commitment aims to ensure that 100% of the energy used by these providers comes from renewable sources.

Signing this Climate Commitment entails a shift in the level of commitment these companies have to sustainability, transitioning from its promotion as a desirable element in 2023 to a requirement for the year 2024. This change in trend is accompanied by a wide range of strategies, including the development of new processors aimed at reducing resource waste, the promotion of equipment recycling practices, and the use of renewable energy sources. Additionally, this shift in trend also brings a significant reduction in the cost of using services, benefiting those companies that adopt a sustainable work model. Furthermore, in addition to the efforts made by cloud providers, it is also necessary for companies using their services to design application architectures focusing on sustainability and to make responsible use of available resources, avoiding, for example, overprovisioning.

IoT

During 2023 and especially in the last two quarters of the year, at Keepler we have seen a growing interest from companies in the field of IoT (Internet of Things) on the cloud and, more specifically, in the field of IIoT (Industrial Internet of Things). This interest that we have perceived is very much aligned with the effort being made by large hyperscalers such as AWS or Azure to help different companies move their workloads, including data from IoT devices, to the cloud.

For IoT architectures, the cloud can be the best ally as it provides capabilities such as scalability, low-cost storage capacity, improved security measures, massive data processing, integration capacity… which are essential for architectures that combine the physical world (the devices installed in the plants) and the world of data. In addition, the cloud allows us to define ephemeral environments in which to carry out proofs of concept and testing of different devices, without representing an unaffordable economic cost for companies, which in the medium term will result in the quality, robustness and reliability of the architectures deployed in productive environments.

In the IIoT field, the migration of workloads into the cloud makes it easier for companies with heavier industrial processes that currently have less presence in the cloud (manufacturing, energy, agriculture, healthcare, etc.) to test and migrate these workloads, allowing them to explore the competitive advantages of having a cloud environment, without dispensing with industrial standards in terms of security, protocols or asset management.

Given the growing interest that we at Keepler are experiencing during the end of 2023 in IoT, we expect that during 2024 this trend will consolidate and the presence of companies in the public cloud will be greater, enriching the ecosystem of data and IoT architectures that we are designing and implementing.

RealTime Archs

The expected boom in IoT solutions in 2024 will lead to real-time architectures gaining importance. Until now, it has been difficult to justify the need for a real-time architecture, both for data ingestion and consumption, due to the higher cost compared to batch solutions. Only products related to trading, monitoring or anomaly detection, logistics or fleet management justified their use.
Are we on the verge of Kappa architectures becoming more important?

In addition to having specific endpoints for the ingestion of IoT devices for proprietary protocols (MQTT, AMQP, XMPP, DDS, etc.), the first challenge our real-time architectures must face is having an event bus scaled to support the entire data volume. It will be essential to evaluate whether managed solutions from hyperscalers (AWS Kinesis, Azure Event Hubs, or GCP PubSub) are cost-effective to scale, or whether we should opt for other solutions like Kafka.

Real-time data consumption will be another challenge we will face when implementing real-time architectures. In this regard, hyperscalers have launched some innovations that facilitate this task: from lakehouses with HUDI, Iceberg, or Deltalake that enable ACID capabilities in the data lake (known as Transactional DataLake) allowing the update of data as it arrives, to services for time series databases.

It’s very important to be prepared for the impact on costs associated with real-time architectures, which involve a larger volume of data being ingested, consumed, and processed. To avoid any surprises in the bill, we must design our real-time architectures with a calculator at hand.

Security and Privacy

The year 2024 looms as a crucial moment for cloud services, especially regarding security and privacy. As businesses increasingly turn to the cloud for their storage and data processing needs, the demand for robust solutions to safeguard sensitive information has never been more critical.

• Technological Innovation:
  One of the defining aspects of 2024 will be the quest for technological advances aimed at strengthening the security infrastructure of cloud services. Advances in encryption methodologies, more sophisticated access controls, and innovative threat detection mechanisms are expected to emerge. The goal is clear: to create an impregnable fortress around the data hosted in the cloud.
• Legislative Developments:
  We will witness a significant focus on legislative solutions for privacy issues. Governments worldwide recognize the need for comprehensive data protection laws. As these frameworks evolve, businesses will need to adapt, ensuring that their practices align with emerging legal standards. The cloud is no longer a lawless frontier; it is beginning to be subject to a regulatory landscape that demands accountability and transparency.
• Enterprise-Client Trust Dynamics:
  Trust is the currency of the digital era, and companies are aware of its role in customer relationships. In 2024, the emphasis on maintaining data privacy is not just about compliance with regulations; it’s about fostering trust. Companies that prioritize and communicate their commitment to protecting customer data in the cloud will not only mitigate risks but also strengthen their brand reputation.
• Importance of Data Protection and Privacy:
  The year 2024 marks a paradigm shift where data protection and privacy take center stage. Companies will be compelled to integrate privacy measures into their business strategies, ensuring that every decision, from infrastructure design to application development, prioritizes user data security and privacy.

In conclusion, the year 2024 promises to be a pivotal moment for the security and privacy of cloud services. Companies navigating this landscape with a proactive approach, embracing technological advances, and adapting to evolving regulatory frameworks, will not only secure their digital assets but also build the trust that is indispensable in our interconnected world.

Where the major Cloud providers are headed

In recent years, the widespread use of cloud services has become a reality around the world, with AWS, Azure and GCP as the leading providers of public cloud services. Further growth in the use of these services is anticipated by 2024, with the growth associated with each of these providers expected to vary according to their particular circumstances.

Historically, AWS has maintained its dominant position in the public cloud market, and this trend is not expected to change through 2024. However, its growth rate will be moderate compared to its main competitors, mainly due to its market share which already stands at approximately 32%.. This situation complicates the possibility of maintaining such high growth rates. For the current year, AWS will continue to focus on the sustainability and security of its cloud platform, in addition to maintaining the strong investment already made in Generative AI.

Microsoft Azure has significantly increased its investment in recent years to try to close the gap with AWS. These resources have been primarily aimed at improving its artificial intelligence (AI) and machine learning (ML)-related service offerings, including the acquisition of OpenAI, the company that created ChatGPT. Azure expects this investment to help further close the gap with AWS, which currently stands at around 10%.

Over the last 3 years, GCP has been the fastest growing cloud platform in every quarter, with year on year growth rates of over 30%. However, in Q3 of 2023, Azure’s growth rate outpaced GCP for the first time. While this could be seen as a cause for concern, it is important to remember that GCP still had a much higher growth than the overall market, and their CEO Sundar Pichai explained that the main reason for the slowdown has been Google’s efforts to help customers optimize their cloud spend. This has reduced short-term growth, but this proactive vision towards optimization and sustainability should increase customer loyalty in the long-term, and Pichai adds that by now, more than 60% of the world’s 1,000 largest companies are Google Cloud customers.

In addition, there are two other trends that paint a bright future for GCP. The first one is the increasing importance of Data-Centric strategies in customers around the world. With Vertex AI, GCP has the most robust unified platform to analyze structured and unstructured data. With competitors still not having fully mature platforms and an increasing demand for customized Generative AI solutions, Vertex AI could see increased adoption in 2024.

Finally, it is worth mentioning that GCP still managed to grow despite being on the backfoot with its GenAI releases in 2023. Azure took advantage of this by moving earlier, but Google has finally catched up, releasing AI integration across its entire Workspace suite with Duet AI, as well as announcing the powerful Gemini. While Azure may still have some early-mover advantage, Google has historically been the leading ML and AI company, and their accumulated expertise, combined with their other services such as Youtube, Search or Google Maps, means their access to training data and potential applications of GenAI is unparalleled.

>>Download the full document Tech4biz Trend 2024<<

Image: Freepik | Rawpixel