TCS Enterprise Agile - Largest Transformation to Business Agility

The session is about the transformation of TCS - one of the world’s largest IT service provider’s talent and cultural transformation into an Enterprise Agile Organization thus providing best-in-class DevOps services to 550+ customers.

MM

Mohammed Musthafa Soukath Ali

Head - TCS Strategic Initiative, Agile Practitioner, Author, TCS

MB

Manikandan Balasubramanian

Enterprise Agile Coach, TATA Consultancy Services

MT

Muhammad Tabrez

Agile Ninja Coach, DevOps Consultant, TATA Consultancy Services

Transcript

00:00:00

<silence>

00:00:14

Hi, I am Mohammed Mustafa. I'm an author and agile coach. I lead the Agile and DevOps transformation for TCS. Today, we are going talk about it today. I have my colleagues money and Tabriz with me. We are TCS. We are a very successful, uh, company. We offer IT services and business solutions for many customers. It is a successful company, and you can see from the revenue graph of pieces for last two that it delivered outstanding capacity increment only year on year. And we are a large company with half a million employees on its payroll. And how TCS is very successful, it always scans the horizon for any change and be prepared to embrace the change. In 2017, PCSR, agile and DevOps pull from the industry. It knew that it had to be prepared, but there were three specific, uh, problems. Number one is thesis had legacy roads, almost 140 number of legacy projects, only 10% A was agile. Rest of them there, the massive portion is knowledge. It's internal processes. Though their agility was better than the competition, it was not matching with its growth aspiration. So this is further researched on Agile and found out that most of the organization in the industry, they applied Agile and DevOps within the IT department, especially for application development projects, not even for maintenance. It knew that the potential for Agile and DevOps is not at a project level beyond to it, and even beyond to the enterprise, uh, level.

00:02:17

Uh, SFAR, hold on. Uh, what is Enterprise Agile?

00:02:20

Mustafa? So, uh, our CEO declared that we have to be an enterprise agile company in 2017, and we had the same question, uh, money organization had that what it meant us as a company, whatever we do, not just the IT alone, not just the IT project alone, but the principles of Agile and DevOps must be applied. Their philosophy is so amicable to any work that we do, right? From marketing to sales research to production and compliance and, uh, whatnot. That is enterprise. So that's a bar that we set.

00:03:00

So considering the mustaf all, we have half a million, uh, employees. Uh, was it even possible?

00:03:06

Yeah. Uh, so it was just to a portion of it, uh, tabs. It was not a joke. Half a million employees, we had upgrade them. We also had to upgrade our infrastructure from non-collaborative, non DevOps, non workplace to, uh, very, very, you know, DevOps facilitating workplace. We have to transform our service delivery projects. Remember, 10% only <inaudible>. So doing all of this within three, uh, years was definitely a challenge. But we went at it. But the challenge, uh, at that point was we have this vision and the, uh, uh, dimensions. How do we move forward? There were two choices for one is the world way of working, phase G, delivering to vanity mission. The second one is DevOps way of, uh, working, eat our own medicine, right? Uh, apply it to agile initiative, DevOps initiative itself, flow feedback, continual experiment. So we have to take a decision here.

00:04:11

So what, what does vanity measures here mean? So

00:04:15

Let, let's assume that we have taken the world way of working to do this work. We would've taken all of these dimensions together, right? Workforce, workplace, all of them together, and we would've done a new huge data crunching analysis across the company. We would've done a huge planning for all of them together. And it'll move through those phase gates, analysis, planning, design, and every stage when we cross, we will mark that we are 23 percentage complete. But realistically, we have not created a single value for anybody. Nobody has consumed our outcome. That is still an analysis, design fees. So delivering to this 23% complete, 50% complete is a r measure <inaudible>. But we chose that there was way of working where we looked at, uh, our appetite workforce or place, and then we looked at what is that minimal thing that we can do, so that can be end to end and it can create value.

00:05:13

So we narrowed down to an aspect called, uh, role transformation, because we had upgrade, remember one 14 roads. And within that, we looked at the most threaten road, which was manual testers. We had 7,000 manual testers on our payroll. We knew that they will not <inaudible> in the digital agile DevOps world. And, uh, where are they? Most of them are in North American region, and they're in the service unit assurance. So go to smallest level, pick up this non-value adding role, and then work on it to upgrade, right? So the first part is flow. Second part is feedback. Apply them, upgrade them on the ground and apply them to project and see the improvement there. Um, the listen to the feedback. Then eventually what kinda value they're bringing back, right? Uh, is the, uh, customer satisfaction improving? Are they doing better than how they were, uh, doing earlier? So that's a continual, uh, experimentation and the pivot. So overall, uh, we were able to do this minimal product of upgrading our manual testers to quality engineers. Just within four months. We have more them up, uh, near career, as well as what tcs to be, uh, future foot in terms of its roles.

00:06:31

Uh, Mustafa, what are the other MVPs? Uh, rebate.

00:06:35

This gave a lot of confidence to us. Uh, money. The first MEB, then, uh, we knew that there were other roles that we had to work on. So we took up, uh, roles such as project managers, uh, middleware, uh, developers, and upgraded them to a, uh, modern future fit, uh, roles. Those were our, uh, subsequent, uh, s

00:06:53

Uh, Mustafa. Every organization does roles transformation, but it not necessarily means maturity in agile way of working. Culture change is critical.

00:07:04

This was a huge learning for us too. Uh, we thought that the role upgrade will, uh, bring agile behavior. <laugh>, it didn't, uh, example is that quality engineers were thinking that they will collaborate with the developers and they'll prevent the defect. Uh, but they didn't, right? They were still thinking the, as a separate QA team, they're waiting for a development to complete capture the defect and so on and so forth. So we knew they had to be jolted. Their behavior should be changed. So we, we brought games, right? Agile games, DevOps games. Many people know. We also send them to external certifications, but none of them work. Then finally, we, uh, found out our own external vehicle called Living Agile, where every employee has to take two days, three days real project and, uh, work on it to create value. If they didn't have agile behavior, DevOps behavior, they'll not be able to create value.

00:07:58

So they'll be told about an assessment. So if you look at the entire, uh, process, right? Uh, first we assess them on their current behavior, for example. You know, are they still bringing the whole hierarchy based mentality, et cetera. Uh, then, uh, we, uh, allow them to take a two day, three day project in the project. Coaches facilitate them to see the non j non DevOps behavior, and they retrospect, they, they change it. So the end of the three day, most of them actually understand and <inaudible> that, uh, behavior so that it'll come out later. Suppose they were, uh, not getting that behavior once second, they have to repeat the project. One example of a project that we have done is for one Indian state called West Bengal, where, uh, the government asked for US beach safety app. Plenty of people who visited the beach, sometimes they get into hazard, and they had let the guard know.

00:08:53

So they wanted a safety app. So we let the employees through a living Agile three day project to develop, uh, the, uh, mobile app safety app. On the third day morning, the team actually rolled it out in front of the offic. So that's an example of real project. It's not a game. It's not just a training. So real time, and most time, most times, you may fail also if your behavior is not changing. So that's a fantastic mechanism. So overall, if you notice now, uh, we have this, um, compet roles as well as we have the agile trained, uh, uh, associates with that actual agile mindset. So we are able to move from a 114 legacy roles company to a seven roles company, and we created a world largest workforce in that process. They improve the productivity for year, but yeah, noticeable, uh, margin. That's why we have, uh, done, uh, respect to that, uh, people part.

00:09:55

But still, uh, most all, uh, half a million employees, uh, cannot sit in the same room, right? They have to be distributed. Yeah. Was that a challenge?

00:10:03

Yeah. PCS, uh, by its, uh, business model is very distributed. Many of our customers are. So when we, uh, applied, uh, agile DevOps initially, uh, uh, no. If you are distributed between, let's say India and us, uh, for you to participate in the daily Scrum, if you are, uh, in, uh, different time zone, then you had to come in a graveyard to shift, right? Non-office and participate. Initially, they did it out of interest, but it didn't last rubber band effect, they came back, uh, to the, you know, I'm not going to make up, right? I'm not come in the graveyard shift. So we knew location was a constraint. Uh, we had to figure out, uh, alternative to Agiles ask for face-to-face communication. So we came up, uh, with a model called 3 4, 5, 3, uh, models, four enablers, five principles. See, the, first of all, you had to find out which time zone that your team is distributed.

00:10:56

If it is one hour, uh, like, you know, India and Singapore, then probably you are, uh, uh, em one model. Uh, you need some enablers. But if you're on opposite side of the geography, six hours plus times of difference, you are M three model, for example, uh, between APAC and, uh, Europe, right? Large time zone, uh, difference. And you need four enablers. For <inaudible> request two, Y four, M three, request four. One example of enablers product, uh, want roll. Uh, if you are in the opposite time zone, the other time zone people may need the duplication of the product, wanna roll. So you have to create a product specialist. You have to create self provisioning infrastructure so that every member in a project will feel like first class citizen. They can, uh, trigger the build. They can set up the data by themselves so that it's on cloud, it is available, uh, not just like the self provision. So these are some of the enablers. Overall, we are able to improve the productivity of our, uh, uh, location dependency teams, as good as co-located teams. Some cases it went even, uh, beyond that,

00:12:03

Uh, mufa digital routine sounds interesting. I understand different digital tools to enable this. So what kind of investments you made?

00:12:11

So we, uh, uh, the digital routine is part of the collaboration infrastructure money. Um, see, uh, for you to be location independent, you'll be able to log in from the device of your choice, from the place of your choice on the time zone of your choice. So the answer was give them collaborative infrastructure. And the branded name teachers had for that was work on agile collaborative workspace, which allows people to, into digital routine. So you have this, uh, conversational tools between the employees. They have the working model set up so that they know when to join and they can sense their whereabouts. So the digital routine for them is set up very nicely through the <inaudible> collaborative, uh, workspace

00:12:56

Place of your choice, uh, I mean device of your choice. How is it, uh, possible, Mr.

00:13:01

Vo? So, uh, you know, a lot of them are, uh, ip, uh, uh, fabric. So I can't reveal much, but I would say this. So we gave employees, uh, seven options to do that right from the place of your choice, us of your choice to securely, uh, reliably connect to the workspace. So this actually came in very handy, uh, later. So overall, if you look at it now, we were, we enabled the people with skills, roles, uh, methods and tools. Now, we have thousands of projects across the company. Uh, thesis is A-C-M-M-A, uh, compliant company, which means we need to maintain the quality baseline. Uh, we should know where you are in terms of your maturity. So we picked up 21 high impact practices. There are around 150 plus Agile DevOps practices. Uh, we can apply all of them, but, uh, you know, to be very practical, we are picked up those that are having high return investment.

00:13:56

We put them in three buckets. Basic, uh, standard, advanced basic means you are to come together in three days. Like daily. Standard, standard means you have to, I create and then demo value every integration that's standard, right? So we, we sort of gave a template to the people, uh, to the project teams, so they're able to assess themselves and then, uh, move from one stage in next stage. For a large, ordinary organization, this kind of guidelines, very, very necessary. You had to make the definition black and white and then get them across. So we, uh, moved many of, uh, these projects, uh, uh, from the basic maturity to advanced, uh, maturity.

00:14:34

Uh, most offa initially, you said 10% of the projects were agile in 2017. So how many, uh, projects as in percentage of projects you moved into agile now,

00:14:44

2017, 1,700 money. We are talking about 2019 now, right now, 6,000, uh, projects we have moved. And these are not just mechanical, uh, general practices supplying, uh, projects. They brought lot of, uh, business outcome, uh, as well. One example here is the first one, uh, on the screen, uh, that is a capital goods, an acute electronic equipment manufacturer. Their shipment and paging got improved because they brought in DevOps, uh, feedback flow and continual improvement, along with other agile practices. So what was, uh, you know, less than satisfactory fulfillment rate, 68 percentage got improved, almost close to a hundred percentage, uh, when they applied this. So we got documented 3,400 business outcome for many of these projects. Uh, we also conduct, uh, <inaudible> conference location dependent, virtual agile conference. So this is from 2019 <INAUDIBLE> conference, where you can see that s employees and customers brought back application of DevOps and Agile on multiple types of situations as well.

00:15:50

So, uh, you talked about, uh, people process. First of all, what about, uh, technology?

00:15:54

Technology we already built in, uh, uh, w when I talked about the open agile collaborative workspace, uh, as well as, uh, this self provisioning in the location. In agile model, self provisioning means, uh, it'll require you to have your infrastructure on the cloud. It'll require applications to be self-contained in terms of, uh, microservices and a PA. So that is one part, right? All that, uh, technical practices and, uh, infrastructure. We also gave them conversational, uh, tools for collaboration that is also technical. But we, uh, see, we're still stuck with 6,000 projects, right? Still, thousands of projects are not legit. When we reali, when we researched them, we found out that a company like DCS are a similar company. It still has this mammoth work, uh, done in a space called, uh, operations. You keep doing the same thing again and again, repeatedly application maintenance, our business operations, right back office work.

00:16:52

So that we found out that a lot of manual hours of spend times, so we introduced Agile, we told that, you know, uh, previous manual should be spent on improving the way that you do, then doing the work, right? Doing the work should be done by the machine. Machine should be given a first rate of <inaudible> on that. But the answer was, we are very busy. We don't have time. So we had to help them out by providing a dual operating model. We call that as a cognitive <inaudible>. Uh, the dual operating model. Part of your, uh, day will be spent on, uh, the doing the work repeatedly. You are still doing it, uh, by humans. It's okay, but you step back for a portion of your time, right? One hour every day, for example. And vaccine is the value through machines, right? You think about what are the projects that I can run to improve the technology agility of the work that you do.

00:17:41

Bring automation, uh, large company like this, I told we had to defend everything black and, uh, white. It cannot be abstract. Uh, people have to be told where there, where they have to go, and they have to be, provide tools also in the scale. So it is still abstract, right? So we brought in, uh, specificity there. We defined a roadmap, uh, from basic standard, advanced best in class. So people of <inaudible>, I'm doing this work repeatedly operations type. That means that you are not using, you are not having any technology agility. Uh, right? You may wanna start with the mirror automation, probably automation of your test cases, automation of your emailings. Then, uh, you go for a purpose driven automation, end to end, uh, processing, automate. That would be the standard. And, uh, we also get a tool, uh, where, uh, it'll, uh, scan your project information, customer feedback, your performance, and suggest you are at this stage currently, and you can go to the next stage if you do 1, 2, 3.

00:18:38

So it became a very, uh, specific exercise. And this was not just a mechanics, right? Uh, cognitive DataOps, we saw very tall close correlation between their maturity to their business, uh, outcome or types of outcome. If they only applied point automation, they were looking at team level outcome, uh, right in sprint automation, <inaudible> ratio, effect containment, so on, so forth. But who applied the best in class, highest level of cognitive <inaudible> maturity. They were able to see organization-wide, uh, outcome, uh, NPS improvement, uh, sales, uh, turnover, innovation, uh, portion, so on and so forth. So that's the technology agility that we brought, uh, uh, to the, uh, place, uh, the base.

00:19:25

So, uh, Mustafa, uh, you have done people transformation, uh, largest agile workforce, and you have done workplace transformation projects transformed, um, agile, uh, DevSecOps been brought in. So you are, a transformation is done, right?

00:19:41

So I wish, uh, it was like that <laugh>, it was massive transformation money. So we still have this, uh, IT services done, business solutions done, but then we still have this huge, uh, part of the organization enablement function. For example, the people who are building the facility maintain the facilities for thesis. They came and told, uh, see for software, we can understand the concept of MEP, minimal able product. We are building a facility, we are maintaining it. What does it mean for us? Uh, people like sales. They told this, uh, dedicated, uh, cross-functional cell sample team. You want dedication, we cannot dedicate. That's not a practical model for us. At one point in time, we'll be, we'll be, uh, uh, sort of pursuing different leads. So for each of them, we have to figure out, uh, what is agile for right end of that, we are able to create a largest collection of agile. We working for each of them with the documented f aq, we call it framework as a agile for everything. This agile for everything converted TCS into a fantastic agile company. Overall, the value stream from the research, uh, till, uh, you know, the, uh, retirement of our assets sunsetting all became, uh, agile. This is a complex, uh, slide. So let's, let's move on.

00:20:57

No, no. Most of all, hold on. Just you yourself mentioned that it's a complex slide. You can just explain, uh,

00:21:04

I think the first one, uh, the first one is, uh, the sense value research. So t research and innovation, before the transformation, they would take, uh, many months, years to file a patent. By applying agile, they were able to reduce it 10 months, 2019, as we speak, I know of many efforts, uh, where we applied patent within four months. That's an example. Similarly, the way that we hire from college, we have increased our, uh, reach and the cycle aim of onboarding an employee. So overall, if you notice that agile for everything, uh, brought internal agility, uh, for example, uh, an associate, uh, uh, who gets, uh, hired in the college, uh, they'll be trained when they're in the college. You have shift after their training back into the college, which means when they join TCS, they're job ready. On the day they join, they'll have a laptop ready on the day they join the project will be mad. So it, it is actually the flow that happens ever since they got onboarded into, uh, TCS overall, this internal agile guests allow the TCS to, uh, help, uh, to transition to a working from home, uh, set up just within, uh, you know, few weeks, half a million employees, their customers, all the work, including emission critical work, they all transform because of the internal agility.

00:22:23

I now understand most of what the seven options you mentioned, particularly

00:22:27

Connecting

00:22:27

Earlier

00:22:28

Place of choice and device of choice. Yes.

00:22:30

Yeah, that's good. Yeah.

00:22:31

Uh, Mustafa Agile for everything is interesting. And, uh, the internal agility which you achieve. Uh, did you take, uh, agile or beyond the company?

00:22:42

<laugh>, it's a dangerous thing to say agile for everything, but we did that, right? Many of us apply agile for <inaudible> personal things, families and whatnot. Uh, in the, we showcase, uh, where, uh, thesis associates took adopted an e Italy, a restaurant in, uh, Noida Delhi, uh, which was having lot of potential, but the customer base was not so great. Sales was not so great. So TCS team helped them to follow Agile we of working to improve, uh, their status quo. Uh, they conducted a daily scrum. They set up listening posts for a customer. They acted on the, uh, the, uh, feedback. Uh, they improved the flow of value to customer, right? So within few weeks, their sales, uh, improved by 23% percent. This is just one, uh, case. We have done many such, uh, uh, work, uh, for a social work. Of course, agile for everything was applied for, uh, managing outside, uh, the company, too many. So how,

00:23:38

How did you do the change management? I mean, any, any specific framework being followed?

00:23:45

It was a problem, right? Uh, TC is a company within a company, complex, diverse. Uh, we tried many things like John co model, many of them, but we had to figure out our own way. We call that as a neighborhood model. I'll explain that. Traditionally, uh, you'll, uh, form a transformation team at the top, and there'll be levels, uh, we need, right? So you will, uh, pass on, uh, the information and the change direction to them. They'll deploy some that sound very hierarchical against the self-empowerment principle that we teach. So we, uh, went with a network, uh, model in which neighborhood will be em embedded. Uh, so each teaches unit, we formed a self-contained transformation unit, uh, as a self-contained transformation team within the unit. And they were connected. Uh, if they are the nodes, they were connected to action centers, uh, called, uh, neighborhoods. Each neighborhood is an expert center, uh, for a specific, specific, uh, vertical location, neighborhood, right? So the whole, uh, information flow, knowledge flow, deployment flow got decentralized. So any point in time, there are 14 neighborhoods that disseminate the requirement, they disseminate the guidance, and then the whole democratization of the flow allowed us to scale a lot better.

00:25:04

Uh, Mustafa 14 neighborhoods. Can you please throw some light on it?

00:25:08

Yeah. So as I told, uh, each of the neighborhood is, uh, staffed, uh, by people who are on the ground working for the customer, but they're experts. There is a cops, right? There is a neighborhood, uh, who is topnotch, uh, talent on this field. They would also be part of the DevOps neighborhood of tcs, and I might be just a junior in s right? Working in a project. And, uh, I know of the maturity that I have to go through from basic to best in class. I have question. I want help. I can strike away in a touch base with this case. There is a central mechanism. It's more of a distributed mechanism. So as a junior, I get the top-notch talent access immediately. Similarly, there are other 13 neighborhoods, uh, that, uh, do a fantastic work of democratizing the information and know pieces, making this as a fantastic, uh, scaling out, uh, model.

00:25:55

Most all having a half a million, uh, employees. Uh, I mean, how do you measure the overall progress?

00:26:04

It was an ask, uh, tab, very light, uh, overall, what do you mean, babe? Where are you? Uh, we had to answer the board. We had to answer the analyst. We had to answer TC, chief, uh, um, offices. So overall, we need to know where you are. So we found out that after multiplication, of course, um, my method called a measurement called agility debt, right? Uh, it actually indicates burden a company carries in terms of agility, heavier burden, your agent debt will be actually poor. It is similar to a measurement like a blood pressure, right? You know what is ideal and very hard to go. So here, ideal measurement of, uh, agile, it is zero. You should not carry any burden at all. If we carry an analogy or force, it is a burden, right? So we calculate this overall number at a project level, looking at the people, process, infrastructure, technical agility. Together, we get a single number. It can be aggregated all the way to intake thesis, uh, as you speak. Thesis, uh, has 0.2 for agility, uh, date. And you can also see that can be decomposed into business units all the way down to project, right? Any point in time anybody can have easy conversation about where you are in terms of your agility. This is not, uh, uh, measurement to scare people away, uh, force the complaints, but at least makes the real, uh, situation on the ground very transparent so you can act on it, right?

00:27:22

Uh, Mustafa 0.24, agility date. Um, what does it mean in a layman term?

00:27:27

Yeah. Uh, very simple to understand. Money 0.24, uh, minus 1.7 6, 76 percentage of the work that we do in PCSE, lesser better. Point two four, zero hundred percentage of the work that we do is agile. So it is a measurement like that. Uh, money. So moving on, uh, the agile data also compared with the growth of our customer, right? Who enjoyed a better agile data, right? Close to 0.1, 0.2, et cetera. They, their, uh, growth for the most part was better. Who is at the lower end of the spectrum, uh, they were not growing better. Uh, many of them are traditional companies too because of some, uh, you know, I guess reasons they were able to survey, but otherwise, uh, they had problem. And we also looked at, uh, the feedback from insight. We collect the customer satisfaction index on the project. We noticed year on year for four years. The customer satisfaction on agile project is always one <inaudible>. Then they have non agile counterpart. That also reinforced our belief that agile and diverse practices really work further business agility and customer satisfaction.

00:28:38

So it makes sense. sfa, I really understand the overall transformation. I mean, you got the one lineer from the CEO, no one Lineer of course, and then <laugh>, uh, then you use the DevOps, you of working for the n entire transformation. And I hope you are getting some good recognitions.

00:28:56

Absolutely. So we got a lot of recognition. Uh, we were placed into leadership quadrant by many analysts. Uh, we got few patents. Some of them are direct. Uh, grant, we also won many awards. So, so the summary of the stories, uh, lot of, uh, legacy roles. That was a company 2017. But when we ended multiple thousands of agile teams with seven, uh, lean, agile, DevOps legacy portfolio of 10 percentage agile only projects that we waterfall, right? Revenue has improved four times now. Uh, so that, uh, you know, we already mentioned, right, 76% is a percentage of the work that we do, <inaudible> and the internal processes also, we were able to improve the overall business objective. But the journey is not gonna stop. So I know <inaudible> are thinking that, uh, you know, we are done with our work. We are not definitely done with our work as we speak.

00:29:51

Uh, we are doing a huge operating model change at a thesis level, but perhaps in another session we can explain that this is about how do you move from a current, uh, geography, vertical based structure to a customer curator journey of incubation, growth and transform. Uh, this is a huge topic by itself, but the point is that thesis is very committed to Agilent DevOps. We have tasted it, our customers have tasted it. We'll continue to refine it. Now it takes us, uh, to the point about, you know, what help that we need, uh, see what we have done so far is a great work. We are very, very, uh, proud of it, all of us. And you wanna share it, right? If you are in a similar stage, uh, you want some help, we can share this knowledge. The second one is, we are also trying to specialize in some of the, uh, areas where the solutions are not that completely yet. For example, how do you measure business? Uh, value? How do you write, uh, agile, uh, contract, right? So there are, uh, Mary asked where we are researching on if somebody's, uh, doing some great work here, they are one stage we would like to learn from them. So that's what we wanna ask. Alright? So thank you so much. Thanks, uh, Manian w for accompanying, uh, uh, the stock. Uh, thank you all. Thank you. Thank

00:31:01

You everyone. Thank you.