The Rise of Cloud-Native SaaS: Key Challenges and Opportunities

The landscape of software distribution and management has seen a remarkable evolution over the past decade, with Software as a Service (SaaS) emerging as the predominant model. In this podcast, we delve into the rise of cloud-native SaaS, exploring the key challenges and opportunities it presents. From traditional on-premises solutions to the revolutionary OmniStrate platform, we examine the transformative journey towards simplified software deployment and management. 

Join us as we navigate through insightful discussions with industry expert Kamal Gupta who is the Founder & CEO of Omnistrate with a mission to democratize control planes and enable businesses, big or small, to build and operate their SaaS in minutes at a fraction of the cost, uncovering the intricacies of SaaS adoption, the integration of AI and ML technologies, data security concerns, and emerging trends reshaping the data community.

AIM: What is the concept and functionality of OmniStrate?

“Building SaaS has been a hard and time-consuming process, and we are here to make it easy for everyone to build their SaaS.”

Kamal Gupta: I think one of the things we’ve noticed over the years is the need for a way to distribute and monetize software quickly. Traditionally, there have been on-prem solutions, licensing solutions, or other business models. However, in the last 10 years, SaaS has become the de facto standard for distributing your software. Building SaaS has been a hard and time-consuming process, and we are here to make it easy for everyone to build their SaaS. What does it mean to build their SaaS? For example, you have software, let’s say, MySQL, and you want to build a MySQL cloud-managed offering. How do you do that? You will need a way for your customers to deploy, scale, monitor, patch, build, operationalize, and automate the day-to-day tasks associated with it, have visibility, and patch. All these things, and many more, have to be solved. But doing that and building all those things takes a lot of time and energy.

AIM: What are the limitations  of deploying OmniStrate on-premises?

“It usually involves a licensing model, which means that as a service provider, you give the software to the client but have no visibility into how your customers are using it.”

Kamal Gupta: The on-premises model is still in use. However, there are significant challenges with it. For instance, it usually involves a licensing model, which means that as a service provider, you give the software to the client but have no visibility into how your customers are using it. You’re unaware of what aspects they find useful or not, their workload patterns, or the challenges they encounter. In the current era, enriched with artificial intelligence and the latest trends, we aim to provide exceptional experiences to our customers. We strive to anticipate and resolve issues before they arise, thereby delivering continuous value.

SaaS effectively establishes this feedback cycle by offering a fully managed experience. This means that consumers don’t need to worry about the technology the provider offers and can concentrate on their business applications. With a licensing model, there’s the added operational burden of managing the software. This includes everything from running and managing the infrastructure to dealing with support queries and maintaining system updates.

In contrast, SaaS eliminates these concerns, allowing businesses to focus on higher-level abstractions. This enables them to offer increasingly valuable and intelligent services to their customers.

AIM: What limitations or challenges did traditional methods of managing services or functions, now offered by OmniStrate through a cloud-first approach, encounter?

“The shift away from on-premises models indicates a fundamental change in preference towards more flexible and customer-friendly options.”

Kamal Gupta: Open source for example has been around for a long time and serves as a valuable distribution model. It excels in building communities and evangelizing products to a broad audience. Each distribution model, including open source, has its unique value proposition, and companies are not limited to a single approach. For instance, many open source companies have adopted SaaS as an additional model to monetize their offerings and simplify customer adoption and feedback processes.

Open source, on-premises licensing, and SaaS are all distinct distribution models. Within SaaS, there are variations like subscription-based services and pay-as-you-go options, reflecting the evolving nature of these models. Over the last decade, there has been a noticeable trend: new companies rarely consider on-premises solutions. The shift away from on-premises models indicates a fundamental change in preference towards more flexible and customer-friendly options.

AIM: How does cloud-native SaaS, particularly through OmniStrate’s integration, facilitate the implementation of AI and ML in businesses aiming to leverage these technologies, and what are the key benefits of this approach?

“The effectiveness of AI models is directly related to the quality and accessibility of the underlying data.”

Kamal Gupta: Both cloud-native SaaS and AI are indeed emerging trends, each fueling the growth of the other. AI requires robust infrastructure, vast data sets for processing, and the ability to integrate with various services to construct efficient models and address specific use cases. Managing and maintaining infrastructure, along with data storage and retrieval, poses significant challenges. Moreover, integrating with diverse SaaS services is crucial for AI’s success.

If these components were not provided as SaaS services, accessing necessary resources and data would be considerably more complex, hindering the development and deployment of AI models. The effectiveness of AI models is directly related to the quality and accessibility of the underlying data. Hence, SaaS plays a critical role by providing the essential infrastructure and integration capabilities needed for AI to operate on a large scale, thereby forming a foundational pillar for AI’s expansion and application across various domains.

AIM: Does hosting an application entirely on the cloud, especially when incorporating AI, lead to higher costs for startups due to the necessity of storing all data and renting resources on the cloud, or is this assumption incorrect?

“This process often takes years and demands highly skilled engineers who can address all the mentioned pain points, representing a considerable operational, resource, and employee cost.”

Kamal Gupta: This is a great point. To address it in two parts, consider a world without cloud technology. For businesses, the need to procure large data centers was a significant barrier to entry. Cloud providers such as AWS, GCP, and Azure have greatly simplified this process, enabling businesses to start easily and adopt a pay-as-you-go model. This flexibility is a key advantage of Software as a Service (SaaS).

Moving a step further, a primary challenge has been the significant effort required to develop a software into a managed SaaS offering. This process often takes years and demands highly skilled engineers who can address all the mentioned pain points, representing a considerable operational, resource, and employee cost.

At OmniStrate, we aim to reduce this burden and lower the barrier to entry. Unlike traditional methods that could take years, with our fully managed service, deployment can occur in days or even hours without any upfront fees. This approach leverages the power of cloud platforms like AWS, Azure, and GCP, underscoring the transformative impact of such underlying platforms on the industry.

AIM: Companies like AWS offer pay-as-you-go cloud services but not the specific SaaS setup provided by OmniStrate, and how does OmniStrate’s offering differ from these existing services?

“They’re giving you MySQL as a service, Postgres as a service, Kafka as a service, OpenSearch as a service, but they’re not giving you the technology where you can bring in any software that will convert into a fully managed service.”

Kamal Gupta: What’s happening is, if you draw the layer cake, there are infrastructure services: EC2 for storage, EBS for compute, networking, all this. And then, there’s this layer where they’re saying, ‘”Hey, you know what, I’m providing you solutions.” So, they’re giving you MySQL as a service, Postgres as a service, Kafka as a service, OpenSearch as a service, but they’re not giving you the technology where you can bring in any software that will convert into a fully managed service. And that’s what we’re essentially saying; we’re providing that and making it possible for anyone now to bring their software and build their own cloud-managed offering. It works across the cloud, across regions, across environments. You can run it in your account, your customer’s account. You can choose a tenancy model. You don’t have to worry anymore about how to build these cloud-managed offerings and instead focus on your core software and continue to delight your customers with innovation.

AIM: How do the costs and practicalities of using cloud-based “pay as you go” models for AI and ML applications compare to on-premise solutions, particularly for startups and new setups?

“Not having access to such resources is even more costly.”

Kamal Gupta: I think there’s a small nuance there. Specifically, when we talk about AI, it appears to be a two-step process. First, you need your model, and then comes its application for users, for whichever corresponding application where customers are drawing inferences, and you’re providing some value. However, when building the model, I agree with you. You are leveraging the underlying cloud providers and their resources to construct that model. That’s the foundational cost you’ll incur. Yet, this is still considerably more cost-effective. Imagine having to establish your own data center, plus maintain and operate it, not to mention needing services like Opas or platforms like OmniState. To build everything from scratch could take years and an astronomical amount of capital just to get started. Now, that barrier to entry or friction has disappeared. You can simply focus on creating your model, refining it, and then begin earning revenue as you deliver value to your customers.

And it’s also about context in comparison. I agree with you; it’s expensive. I’m not saying it’s not, but compared to what? Not having access to such resources is even more costly. For instance, lacking CloudFlight could cost you ten times more.

AIM: How does Omnistrade ensure data security, and what frameworks are necessary to guarantee the safe use of data?

“There are measures like access control for determining authorization, authentication, and encryption to protect against unauthorized access.”

Kamal Gupta: I think when it comes to security, proper mechanisms need to be in place, along with a platform layer that can enable them. These come in various forms. One area involves a set of controls to reduce the surface area of attack, which can be achieved through defining the network firewall, typically implemented by cloud providers as VPC or similar variants. Then, there are measures like access control for determining authorization, authentication, and encryption to protect against unauthorized access. Additionally, measures for DDoS attack prevention are essential. For example, AWS provides several solutions for these concerns: IAM for authorization, Cognito and GuardDuty for security monitoring, and the concept of VPC for network isolation.

The challenge, however, lies in the complexity of cloud providers today, offering hundreds of services, which can confuse customers about how to leverage or integrate these services effectively. This is where OmniState comes in. We simplify this process by integrating these security measures into a cohesive package. This approach covers everything from reducing the attack surface area to monitoring through auditing, incident reporting, and enhancing observability. We provide these integrated solutions so you don’t have to worry about configuring and orchestrating them yourself.

This underscores my earlier point about the cloud’s role, especially in building managed services and cloud-native applications. Cloud providers offer a wide range of tools and services, which you can select and assemble to create a secure solution. OmniState’s role is to streamline this process, offering a packaged solution that spares you the complexity of piecing together these components yourself.

AIM:What are the emerging sectors embracing cloud-native SaaS solutions, and where do you foresee the greatest potential? Additionally, could you provide insights into notable use cases within these sectors?

“AI itself facilitates the development of SaaS services, creating a feedback loop that fosters the creation of value-generating intelligent services.”

Kamal Gupta: We observe serverless computing, for instance. A considerable effort has been made in the past decade to transition towards serverless computing, aiming to conserve resources and address the cost factor you mentioned. This involves shifting workloads to operate only when necessary, rather than keeping resources running continuously.

Another notable example is AI. With the vast amount of data and infrastructure we handle, there’s a noticeable trend fueling AI’s growth. Moreover, AI itself facilitates the development of SaaS services, creating a feedback loop that fosters the creation of value-generating intelligent services.

Additionally, edge computing represents another significant trend. We’re continually expanding the boundaries, bringing cloud-native SaaS to the realm of edge computing. This is evident in various sectors such as automotive, where innovation is a daily affair, and healthcare, with the implementation of improved EHR systems. Even in retail, the evolution from Amazon’s inception to today’s standards, with concepts like one-day shipping and DoorDash, underscores the impact of SaaS applications.

Indeed, behind the scenes of platforms like Amazon.com, there are numerous microservices operating as SaaS offerings, each contributing to a complex web of solutions that power various sectors on a day-to-day basis. Take banking, for instance. Companies like Razorpay in India are revolutionizing the fintech industry, offering fully managed SaaS experiences that redefine customer interactions.

In summary, while these trends span across industries, the prominent ones include serverless computing, AI, and edge computing, all of which represent significant shifts driven by the cloud.

AIM: What are some of the potential  challenges, opportunities, and key trends within the data community in the near future?

“All of them face the distribution challenge.”

Kamal Gupta: When it comes to data or AI applications on Cloud Native SAS, I believe that examining data and AI applications reveals a significant trend. Over the past decade, the pace at which new data solutions and AI applications are emerging has accelerated tremendously. Almost every other week, a new database is launched, and numerous startups, though I lack the exact count, are entering the AI domain regularly. However, all of them face the distribution challenge.

Reflecting on the period before 2015, the SaaS model wasn’t proven. The first-generation SaaS companies, such as Databricks, Snowflake, and Confluent, have revolutionized the next generation of growth in both the data and AI realms by validating this model.