Join AITSM Co-Op, A community for those who ♥️ AITSM (ITSM, ITIL, AI and Automation for ITSM)

A Productive Rant About Automation (Part-2) - Are AIOps & Automation CoE Mature Enough?

June 14, 2022

A Productive Rant About Automation (Part-2) - Are AIOps & Automation CoE Mature Enough?

A Productive Rant About Automation (Part-2) - Are AIOps & Automation CoE Mature Enough? The Biggest Hurdle To Automation and How To Face It?

A Productive Rant About Automation (Part-2) - Are AIOps & Automation CoE Mature Enough? The Biggest Hurdle To Automation and How To Face It? 

"(On making happen the hardest changes) Never drop the ball. Consistently put in the efforts and always go with the data. It comes, maybe not fast as you want, but the result comes. So have that optimism. Okay. And consistently work and explore the different approaches. Understand the challenges of the person and why there's resistance. And try to see if you have the alternative approach where your ask for enabling the automation and their fear of automation can be addressed. So try to figure it out and have little conversations about it. That's what I feel works. Maybe it takes some time, but it works." - Says Sapna Gupta, our esteemed woman thought leader and guest on the AITSM show. 

Sapna Gupta is Infra Automation Advisor at Global IT Automation CoE, Cargill, having 20+ years of experience. 

She put forth a slice of wit and forthright explanations to uncover the maturity of industry practices. Whether AIOps & Automation CoE Mature Enough? What is the Biggest Hurdle To Automation and How To Face It? How Did She Solve One Intriguing Tech Problem?

We'd learn a lot from her plenty of examples, enthusiastic spirit, and point-by-point conversation from 2+ decades of rich experience. 

Don't miss the end, where she shares three crucial things to boost any automation initiative and a special message with career advice for all women leaders.

"Wherever you are on your journey. Please keep adding the skills, so you are never going out of date. Automation and AI-ML are the latest things, whether you are on the Infra side, the application side, or the developers' side. Please follow the latest trends in this space to help build the most successful and stark career." 




Learn and Grow, With AITSM Show!!!


Anand Tambey: Welcome back to AITSM show. And we are having Sapna Gupta with us. We will explore more insights in this episode.

So how do you see what is the maturity level of current infrastructure automation COE, and like what kind of things you propose, which could actually improvise the COE as well as like IT service management through the use of automation, as well as AI, not just chatbot, but the using the NLP or prediction mechanisms and all. So what is your view on that?

Sapna Gupta: If you see first let's talk about the automation maturity. I will say from the very high level, it's the task-based automation, which used to be like early days in hype if you do write some scripts to automate some of the tasks, itself was a big enough.

But now in the, era what we are we have to take a step ahead. And from the task base, we're actually moving towards the process, end to end the process. Okay. And which can actually improve the client experience. Now there is a workflow based automation. Okay. And then integrated with different IT services. And then the next layer it comes to using the ML and the AI into the automation. So that it's not just a well defined kind of things where we are actually writing the code to define a process or the series of the activities or the tasks. More than that, it can actually identify the problems, apply some of the ML algorithm on top of that and can make a prediction of the probable issues, depending on the historical patterns.

Cargill is an agriculture company. So there are some of these specific months where the certain applications are in use that the agri business months in the entire year. We know that this time the server may have required a high, resources to support that kind of a high volume is coming for the execution.

It is better to have some precautionary steps, maybe add more resources or have a kind of a load balancing architecture enabled. So that, the load is getting distributed without creating the impact on the business users. Okay. And that kind of predictive analysis and the related steps will help actually to provide the seamless experience. Okay. So this is one example of it. Okay.

Now, How we are going ahead on that integrated automation approach is something still I feel we have a very silo mindset. Okay. So if it is an infrastructure person he will just talk about server, the database or the network, if it is an application guy, he just had the visibility of the applications and the same for the middleware.

In the each of the teams automation person for that particular service. And that's how the binding is happening. It's like they understand that particular technology tower, and they also the SPOC for connecting on the automation part with this agile COE. And that's how that, that practices can be for followed.

So this is a people part of it. Another is that on the tooling part of it? Okay. Now most of the times, every team has. separate tools. Okay. To do the monitoring, to do the automation, to manage their own landscape. The IT landscape. Okay. So let's have one platform kind of a thing.

So if you see many of the organization, they are going towards the AIOps. Okay. Which means that it's a single platform which can gather all the logs and the metrics data from the different monitoring tool and create a single view. And through that single view you can understand how this application has the multiple dependencies or the different components.

And where what has got broken and it's coming actually appearing in a single view. And that's a kind of a platform which is enabling the integrations of the different teams. Another example maybe common platforms could be like Ansible or Jenkins to allow their automation scripts to develop. Okay. So centrally we have a kind of a platform which is actually giving a flexibility for the different teams to create their own scripts and all the automation tasks, but all the governance or the own those scripts or the view we are managing centrally.

And also we can use the GitHub repository whichever the tool you use but the outcome of that they're, storing centrally at one place so that we can have that integrated view or it could be a data lake or on different ways actually.

Anand Tambey: Yeah. So you are integrating it on one platform and suggesting like it should be a kind of AI ops or a one single platform with integrated view with every system.

Right. And integration with every system available. Right. So from the AI ops perspective, if we go ahead right now, what is its maturity level so far that AI ops is doing what it should be doing? Or there is something more to be done?

Sapna Gupta: So there are many tools in the market offer a lot of AI ops capability.

And I feel, this is an era where we are integrating all the monitoring alerts and getting in a single view. How we can use the integrated view t o self validate, some of the repetitive issues, which is very common when I will say the space crunch which looks pretty simple issue, but when the outages happens because of that. There was a space crunch. Okay. So how we can actually have the self remediations coupled with AIOps platform, which can help to address 70%- 80% of the issues.

So that engineers, analysts focusing on the rest of the 20% and their efficiency levels will be very high. I think that's the level that we have to reach, but yeah, everything comes with slow steps, but I know that many of the companies are moving towards it.

Anand Tambey: That's correct. If you have a magic wand or something? This needs to be changed in my area. If I, if I can change it.

Sapna Gupta: So I will say, the moment we have some requirement, like it all depends like, what is my role? If I am a service owner, I believe that you took the person, the it engineer or the IT manager. The first thought should be the automation.

If I have some, a roadmap for the next year, I going to do XYZ. Now the automation should be integrated in the plan. How I can ensure this will all the repetitive tasks is actually 100% getting automated. That's what is the target that each of the IT person, we at the different level of the manager level, engineer level, and the L1L2. Okay. All of the people, they have that mindset and that's the magical stick I want. To change the mindset because it's not about the tools or the things. It is more about the mindset.

Anand Tambey: Well, is there a challenge in your two decades of experiences where you have actually successfully solve a difficult problem? So definitely our community would like to hear about like how you have solved that problem or got the success?

Sapna Gupta: Yeah, so I, I will call out one of my experience s which is not the latest one. It was at the time when this configuration management tool was just appearing in the market. And I was supposed to manage a lot of servers.

Okay. And everytime one issue used to come. It's like, whenever we're rebooting the server, all the mappings are getting lost. It was the session variables were getting defined. Most of the times the people were forgetting to make the changes, which make the entry registered after the week.

And that was impacting a lot of like application outages because once the server is up ohh that application is not working it has a dependency on some of the mappings of the network. Now, I got that times some scripts create databases. It was taking the snapshot of those mapping, the session variables.

Okay. I'm storing that to set workspace. So even if server crashed or rebooted if you have to rebuild all server coming up and if something is missing, we have some restoration point to a refer it and then just execute that as a script. So it's not just a mapping is getting stored, but it was to stored in such a way in this such scripted way, we just run it and it will restore all those mappings . we don't have a concept of centralized managing on ansible, chef and puppet was not in the market. We do follow a lot of tools and the major challenge, which we face is with the mindset. Now we grown up in the technology.

 You have the problem, the solution is there. Tech solutions are out there we are into that era now. The main problem comes, what I face is with the mindset like who has certain way of doing from lot many years facilitating the resistance for the adoptions of the new way of doing, maybe a little threat. So that's the major challenge, which I keep on working.

Anand Tambey: Right? Right. I remember one cartoon, which has like, some people are having a square wheel and then they are having difficulties and another person is suggesting a wheel there, but they're saying no, no, we are busy. We don't want it.

Right. So you might have like a lot of things we need to actually make a change. Right. So what are your way of dealing with it when you are going for, innovation and automation and likewise and then propagating it.

Sapna Gupta: Yeah, so I believe, and never drop the ball. Consistently put the efforts and always go with the data. It comes, maybe not that speed as you want, but definitely the result comes. So have that optimism. Okay. And consistently work and exploring the different approaches. Understand the challenges of the person, why there's a resistance. And try to see if you have the alternative approach where your ask for enabling the automations and their fear with automation can be addressed. So try to figure it out and try to have little conversations about it. But that's what I feel works. Maybe it takes some time, but it works.

Anand Tambey: Great, great. Yeah, it's a useful and valuable. So as to conclude or just summarize our discussion. Right? So you have discussed about automation, AI, and like what kind of Challenges we had and kind of what, how we could solve it. So could you just give me three points where you are like not satisfied with whatever, is there in automation or what the three things you want to change?

Or rather, if we can say, what are the three things which could be changed, which can actually give automation industry a boost

Sapna Gupta: yeah. The first thing is the specifically the budget allocation planning. Okay. You cannot do the automation without an investment. Okay. And most of the people they have think if I automate. I should not invest rather I should save. Savings comes later, first the investment comes.

Okay. So that is the one thing that every manager they should have in the mind and allocate the budget to enable the automation. Okay.

Second thing is like the mix of the talents. Okay. If any of the team has the very experienced people. Okay. Who has first very good experience, but more point of the, way of working of like as they were doing 10 years before.

Okay. So like let's have some of the new talent which can enable more broad discussions and mix of the expertise, the prior experience, along with the new technology new approaches. Okay. So that makes it very important.

Third thing, to limit the number of the tools that you want to use? It doesn't mean to restrict only to one or two. Maybe it's a couple of tools, but the moment we want to bring more and more tools for the automation. So we are moving our focus from doing the automation, to the, managing the tools, increase the technical debt with the other sides. I have automated this by introducing this tool another task I automated through this tool and then over the year, maybe you've got maybe hundreds of the tools into your environment. And then it becomes a challenge to manage the tools landscape rather than the IT landscape. Okay. We have to be mindful about it. I think these are the three very important points.

Anand Tambey: Yes, yes. And then these are ideas which can help the community to think about these important things. Thanks for being on our show. We learned many things from your decades of experiences.

And and any, any last advice, if you have, you are in a room full of people, and then there is one, advice you could give.

Sapna Gupta: Yeah. I would like to say for the since it is your women specific episode. Okay. I would like to say to all the women leaders.

Okay. So wherever you are in your journey. Please keep adding the skills so that you are actually never going out of the date. Okay. Automation and AI-ML is definitely is the latest thing be you on the Infra side or the application side or developers side. Okay, please follow all that latest trends in to this space definitely is going to help in building the most successful and the stark career.

 Thank you. Thank you. And thanks a lot, Anand for giving me the opportunity. Yeah, I enjoy talking to you.

Anand Tambey: Yeah. Yeah. Same here. Thanks. Thanks a lot.

Thanks a lot for listening. Please let us know what topics we could add here . And give us the feedback. And if you have loved it. Please share it generously.

Thanks again. Bye-bye.

Sapna Gupta Profile Photo

Sapna Gupta

Infra Automation Advisor at Global IT Automation CoE, Cargill

As Infra Automation Advisor and member of Global IT Automation CoE, Sapna is accountable for developing automation framework, automation maturity model, strategies and automation roadmap for global IT.