Hi everyone, welcome to Con 42, DevOps Summit 2024. And I'm Sora Panda. And today I'm going to give my thoughts on AI as a catalyst redefining the resource management in DevOps. So let's dive in. So today's agenda looks like this. First we are going to go through a basic introduction of DevOps and evolution of DevOps in the last decade. And then we'll talk about how AI and infrastructure as code can work so well in unison. And then we talk about application of AI in automated optimization, enhancing security and error detection and self healing. We'll have a couple of case studies to look into how AI can actually work in real world at today's so, I'm Saurav, I'm the founder and CEO of Cloud Code AI. Cloud Code AI is building an AI enabled DevOps platform which automates the resource management for you so you as a developer can focus on building your product, adding more features to it, rather than spending time on managing resources on the cloud. We aim to help software development teams go to market faster and rapidly prototype and keep on building their product. Feel free to connect with me on any of the social media platforms and let's dive in. So firstly, what is DevOps? DevOps combines development and IT operations for better collaboration, communication and integration between software development and IT operation teams in big firms. It aims to increase the speed, efficiency and quality of software delivery while reducing the time to market and failure rate of new releases. As you can see in the diagram, DevOps is the union of software development and operation. As a DevOps engineer, you need to write code and also monitor your system and make sure everything is running well. Now this is a diagram which very perfectly illustrates what is DevOps about. In the world of DevOps, collaboration is not just a buzzword, it's the backbone of our success. When development or operation team unite, the magic of innovation and efficiency truly happens. As you can see in the DevOps pipeline, it's an infinite loop of planning, creating, verifying, packaging and releasing your software and thereafter monitoring it. And this process keeps on happening. And using the DevOps process we can ensure like what we are deploying or redefining to the production is ready to go there and there is no problem and customers have a better experience. Now let's go through a brief history of DevOps in past decade. So in 2007, the concept of DevOps emerged to bridge the gap between software development and it operations team. The aim was to create a more integrated and efficient workflow where both teams could work together towards a common goal. By 2010, the term DevOps gained traction, especially on social media platforms like Twitter, where hashtag hash DevOps sparked lively debates and discussion among team. This showed that industry was paying attention to DevOps, and it was more than just a trend. In 2015, DevOps was incorporated into scale agile framework, gaining more traction in the enterprise arena. And by 2016, most of the high performing companies began adopting DevOps as the new norm when deploying software. And by 2019, enterprise began embedding more it functions such as security, privacy policy and data into their DevOps culture and processes. So this is how DevOps has been evolving. But in general, DevOps consists of a lot of automation and involves a steep learning curve to understand the current technologies. And there are a lot of tools and technologies popping up every now and then. Thus, convergence of AI and DevOps is a game changer and let's see how it can change lot of aspects of DevOps. Now the first thing is infrastructure as code. So what is infrastructure as code? Infrastructure as code is a process of automating, provisioning and configuring infrastructure using code and script. It helps you increase the speed and efficiency of software deployment, along with better documentation, increased scalability, better collaboration, less human error, and ultimately which results in improved customer experience. Now, AI can actually have a great impact when it comes to infrastructure as code. The traditional way of creating infrastructure as code involves manually writing code or relying on predefined templates, which can be time consuming and repetitive. However, AI has revolutionized this process by allowing us to generate IAC code through simple natural queries. This speeds up the process and increases the productivity, allowing engineers to focus more on the complex task of designing the architecture rather than writing the code. Moving on to proactive assistance, AI can optimize infrastructure design, identify areas where costs can be reduced, and enhance overall operational efficiency of the whole DevOps pipeline. With AI's proactive assistance, our infrastructure structures can be robust, cost effective and fine tuned for the specific operational requirement. In the world of DevOps, ensuring quality and security of the pipeline is of utmost importance. AI acts as a vigilant overseer, helping site reliability engineers ensure quality and security of their DevOps pipeline. Now this is a case study of generating serverless API using natural language. Here you just need to pass a natural language query of what you want to do and AI can automatically identify all the resources needed, like DynamoDB, table lambda functions, and API gateway. It can configure the IM policy necessary to connect each of these services internally, and then when you click on deploy, it can automatically create an IAC file which is a terraform file and this is like a 400 line long terraform file generated in minutes. So you can see how quickly AI can help you build things and speed up the efficiency of your DevOps team. Now, automation in DevOps is like autopilot for an aircraft. It doesn't replaces the pilot, but it empowers them to fly higher, faster and with greater precision. AI is not just a tool, but game changer that enables DevOps team to adapt dynamically, predict demands and optimizing resources with minimal human intervention. Infrastructure optimization is crucial in managing IT infrastructure efficiently. Teams can use AI insights in adapting to usage pattern and anticipating demands, and optimizing the infrastructure with minimal human intervention. Teams can also utilize AI for resource optimization by analyzing the historical data, usage pattern and performance metrics to optimize resource allocation and ensure efficient use of infrastructure. One of the major contributions would be in automating testings, where AI can automatically create test costs and pass them through the testing pipeline, helping you improve the speed and accuracy of the testing. Now let's discuss the role of AI in improving security. AI can analyze vast amount of code rapidly, which is a game changer. It can quickly identify potential vulnerabilities that might escape the human scrutiny, enabling us to address these weaknesses before they can be exploited. Moreover, AI spots these issues and suggests the improvement, streamlining the process of fortifying our digital defenses. Moving on to compliance, which is equally critical. Compliance checks are often time consuming and cumbersome, but AI is transforming this narrative by automating both the standard as well as user generated compliance checks. AI speeds up development and ensures a high degree of accuracy in adhering to regulatory standards set by you and the open source community. This automation is particularly valuable in today's fast paced development environments, where speed and compliance must go hand in hand. Lastly, let's dwell into access control and identity management in our interconnected world. Managing who has access to what paramounts it's not only about humans, but it's also about the services and who is accessing which services. So here, AI can step in in revolutionizing IAM. By analyzing the usage pattern, behavior and connection between the services, AI can generate dynamic IAM policies, ensuring that the right people have the right access at the right time. This not only tightens the security, but also enhances the operational efficiency of your whole product. This is a case study which we created an IM policy generator for AWS. Here you can pass a natural language query asking what type of services and what type of permissions you need and AI can automatically create a least privilege policy for you which you can copy and paste and use it in your services. This is just like tip of the iceberg. Imagine once the AI gets the access to all the data of how the things are connected, it can automating, identify and give the limited permission which the services need. Now finally, we are at the last application of AI, which is in error detection and self healing systems. The concept of self healing in technology is a breakthrough in its own sense, but aipowered self healing system will represent a major leap of faith for the DevOps community. These systems are designed to detect and resolve issues as they arise automatically. This means significantly less manual intervention, reducing downtime and enhancing overall reliability of the system. Imagine a scenario where system issues are addressed and resolved even before they are escalated into a major problem. That's kind of proactive maintenance AI brings into the table. The other aspect is predictive analytics. The integration of AI into DevOps monitoring tool isn't just about responding to issues, but it's also about predicting them. Through sophisticated algorithms, AI can analyze patterns and trends to foresee potential system issues. This predictive capacity allows team to be notified, often before problems occur, enabling them to take preemptive action. Lastly, anomaly detection. Here is where AI can truly shine. AI tools are adept at continuously scanning engineering costs and data for any anomalies or irregularities. Upon detection, these tools can immediately initiate corrective actions, often without the need for human intervention. So AI truly has a lot of potential when it comes to error detection and self healing systems. So just to conclude, AI is going to be a big game changer when it comes to it resource management. And there are going to be a lot of AI powered tools coming into the market. But one of the crucial issues which have always been there is like a lot of tools have been popping up in DevOps and all of them mostly have a steep learning curve. And the main goal, main aim of the DevOps community should be in ensuring that the learning curve is brought down. And we use AI to actually empower the team and the DevOps and SRE engineers to actually build more resilient and reliable systems and focus on building and designing infrastructure, rather than doing the manual grunt work which is usually involved when it comes to DevOps. So AI is going to help a lot of DevOps team speed up the efficiency. And you can also see lot of teams building and moving faster in software development cycle. And one more prediction would be like, you can see the cost of software going down pretty soon. So that is to conclude this presentation and AI is truly going to revolutionize how the resource management is done. And we might also see like AI enabled AWS which can be like you don't even say anything and it takes care of everything, like assigning IP, assigning domain name, connecting them and doing like AI software engineer of its own. But I think that is a little bit in future. But there is lot of scope and lot of hope. So thank you so much for listening to me. I'm Saurav Panda. Feel free to reach out to me on on Twitter or LinkedIn or shoot me an email at sort of at the ratecloudcode AI if you have any questions or if you are just excited on what we are working and want to connect with me. Thank you so much and looking forward to hearing other speakers.