 Brazil and Qatar based firms and I happened to internet Microsoft long back For the past few years. I've been working with Nutanix as a designer and one of the products that I was work that I've worked on is IoT probably you can check the link Later on it's a quite interesting enterprise product. It falls into a space of enterprise and consumer grade In my free time I sketch I sketch a lot I paint a lot That's my Instagram handle in case anyone wants to check that out as well Now let's get to the point What is enterprise design if if if you have any Pre-concept notions about it or anyone knows about it Would anyone like to define it? Yeah Yeah, but what do you mean by enterprise products and how are they different from consumer grade is what I'm trying to get So let's There's a scale aspect. Yes, definitely There's a lot of data involved. There's a lot of information involved. Yes, and and they're all correlated So there's this inherent complexity in itself, right? I tried to simplify it as much as possible What enterprise design is is the design of products for people at work. So when consumer is designed for people Enterprises designed for people at work now every time you wake up in the morning You look at your iPhones you look at your iPads you look at your MacBooks and all the other great devices that you have and you you have These need delightful experiences that you wake up with till the point you get to the office, right? And once you get to the office, this is what you get, right? And there is there is absolutely no reason this should happen I mean the enterprise solutions can be delightful. They can be need They can be efficient and at the same time they can be as well-designed as our consumer grade experiences so What I've tried to do here is for the past two years. I was working on an IoT product and again It fell into the enterprise zone I tried to find out things like tips and tricks that I could use to get this enterprise experience as close to the consumer, you know Design as possible now some of these things are extremely obvious Some of these things are common sense some of these things you already know Some of these things could be wrong like these these are my opinions in some places But anyway, like they've helped me a lot in designing in the past and probably in some way if they can help you I'll be really happy about it. So let's go ahead now as we said the first thing that is extremely different in the case of a Enterprise design problem is the domain is alien You never have a clue of what sort of operations you are going to design for Sometimes you're designing for space. Sometimes you're designing for health care Sometimes you're designing for defense and there are a lot of these different domains that you are never aware of So that's that's one thing the domain is always unrelatable. Now. How do you navigate through this problem? How do you deal design for something which is foreign to you? There is no escape to it you have to do the research There is no other way you can come up with an efficient solution without really studying that domain, okay? So this is this is the obvious point Moving on to the next part. The user is different in this case So I am very sure all of you have heard this we need to design for our grandmoms, right? Make a product so simple that even though your grandmom can use it but in enterprise that's different that we are dealing with a different breed of users and Definitely, they are not your grandmoms. So what you should keep in mind is you should always keep in mind the trinity One is the business guy who is going to sign the check Right one is the provisioner the managers who are going to distribute these products across the company And then you have the end user who's going who's like a developer or someone who's actually going to use that product So when you're designing just keep in mind that you need to appeal to business goals Execution goals and then the goals of the end user The user is expert as I mentioned we are not designing for our grandmoms in this case the user Most probably has more knowledge about his domain than you have So the user research somehow like in some places becomes easy You just need to ask you just need to ask the right questions and you'll get the right data, right? So don't like It's it's okay to rely on what the user says in this case Otherwise you have to decipher in consumer grade you have to understand what they really meant but here it's slightly easier But this is another problem like if I if I ask you how did you get here? What would you say? No, no as in how did you get to the conference? I'll make it simple. How did you come to this hotel? You took a cab and you came here, right? So whenever you ask someone how do you do something that tend to simplify that process? Like take I took a cab and I came here, but that's not how it happens, right? You dress up you check if you have money in your pay team and what not if you added your credit card Then you book in you wait for the mover then you call the guy So there's always these invisible steps involved in getting that work done, right? So don't Accept everything that the expert user says you need to investigate a bit So you need to sit beside them and really observe them doing that task Now the biggest problem that I feel in enterprises the access to the user is difficult so like in Nutanix case we have like 60% of the business is from federal clients and these federal clients are Federal bureaus of the Americas and they they have no intention of giving you any information They are going to keep their data as confidential as possible even if they have problems They have their own support. They don't even reach out to your support Support teams so in such cases it's extremely difficult to Talk to the user or get his feedback get the user research done as well So how do we how do we deal with that? So you should find surrogates There's a good chance that around you within your team You have someone whose role is as close I mean close to what you're designing for whom you're designing for So if you look around in your social circle if you look around in your team if you look around in the engineering org of your Organization you'll probably find someone who is you know close to that role and you can ask them and pms Definitely, they they are gods in this case. They are the ones who can help you the most Find symptoms. So because you cannot Really do the user research and get the pain points and see how they are reacting to the product You need to find other avenues to get The feedback so what you should do is you should probably go to the places where people complain Like review channels like places where people take raise tickets like to a support team It could be community pages where people just go and you know bash products that this products this products And stuff like that and you can carefully pick stuff from there and that that's your feedback now And then probably you can get to your desk and work on that The next problem as you correctly mentioned is the scale the scale is huge and really really use like the IOT product That I was designing for a single person on a single instance was handling around two lakh different IOT devices Okay, and that's that's huge around like spread across some five airports So the scale is you so whenever you start a project that has a huge scale you should always define it So let's say you started with something like a Background app for uber you should always define a ballpark figure, which you know, which is your not star now So you say I'm going to design for thousand operations For one lakh users in a single instance and that's your not star now. That's what you designed for and but It's it's it's easy to get intimidated once you start talking about those numbers Right, I want to design for one like people who are going to use this one product at the same time So it's going to get extremely difficult and intimidating intimidating in the beginning What you do is you start small you start designing for 10 and then you stress test it for 100 people if it works for 100 it could be as simple as designing a table So you design a table for 10 items Then you see if scales to 100 and then you keep on adding customizations to you know really scale it out to thousand items And again, I have not put screens because like I am very sure you guys can visualize what I'm talking about Always show insights. So summary is always help since we are talking about one lakh two lakh Objects it always helps to summarize that data So it could be the title of your table which could give me a clear insight of how many items are there How many are filtered how many are okay? How many are not it could be widgets on your dashboard as well that tell me What my entire system's health is about? Okay, so always always whenever possible show insights Allow manipulation when you're dealing with something that is to that scale You should always design power and this these are the things you should do Right up front and not delay for later releases. You should always start with you know designing Very strong manipulation tools like filters and sort sortings and stuff like that always add metadata so now When we are talking about devices device it could be belonging to different families, right? It is difficult to select a bulk of different types of objects Unless you have given ability to tag them So every time whenever there is a creation workflow whenever you're adding something to your systems Make sure that you provide an option to tag it So now once you've tagged things like let's say blue shirt and I tagged every person walking into this hotel Now even if that person is a male female old young doesn't matter if I select blue Shirt tag it's going to automatically select all the people wearing blue shirts here and that makes my job easy later Introduce modularity again. Yeah in terms of application. So let's say your application is built of different things I mean, I'm just building an example on what you said. So let's say your application is built of like scripts and You have your storage somewhere else Now imagine you have hundreds of these you have 100 scripts. You have hundreds storage accounts. You have 100 Like ML models and you've built an application on top of it, right? now while creating that application I can't tag it by saying health care or UX India conference, okay attack that application now all the items below it get tagged with UX India conference Now tomorrow, I have to build like the next year. I have to build an application for UX India again I don't need to scan that entire list and you know find The storage accounts that are assigned for UX I can simply select assign everything that is tagged with UX India to this new application and boom, that's done So that's what I mean by smart tagging that does it answer the question. Yeah Suddenly when technically said Okay, so it's complicated obviously when you have something at such a scale and you have so many operations going around It's bound to get complicated, but I really find it satisfying So I'd like to give an example here. Let's say you walk into a restaurant. Okay, and you sit on the table What is the next thing that happens? Yeah Sorry Yeah, or the the waiter gives you a menu, right the moment you sit down because the waiter understands your intent He understands that you are not here to like play football or something. He knows you're in a restaurant You've sat on a table. So you're going to order. So it gives you the menu. That's mirroring the intent of the user and In all the ways possible in your enterprise solution, you should try and reflect the intent of the user So in Nutanix, we have something like one click upgrade So usually upgrading a particular, let's say software would take you to do 20 steps in In an enterprise world, but on your phones It's it's it's just a click of a button, right because it's understanding your intent of updating that software and it takes you right The system drives it. You don't need to take all the efforts. So we we try to do something Similar in the Nutanix environment We we provided one click. So you just click upgrade and the system figures out the rest So always make it intentful have an opinion. So let's say you you're struggling you're struggling with the menu You're struggling with deciding what you want to eat Now the waiter looks at you struggling and he comes to you and says sir order biryani What he's doing is he's having an opinion about your and like he reads the situation and he has an opinion about it So he's going to say he's going to suggest something that he feels is right for you Right, so even as designers we should have an opinion even if the user is an expert like whenever we are designing dashboards We should have a very strong opinion about what should go on that dashboard because there's so many stats to show and everything is important for the developers But we need to really filter out things that we think are important and obviously it has to be based on an educational guess and research Simplified structure. So a lot of enterprise Products, especially the legacy ones. They are they have a problem with this So you you have a side navigation bar. You have a top navigation bar and there's their nested Menos and stuff like that. What this does is it it makes it extremely complicated to way find So whenever you're designing a enterprise solution Trans tick to a single type of navigation So if it's sidebar then sidebar if it's top navigation top navigation stuff like if it's search only then search only But try and simplify it introduce some hierarchy. Not everything is a first-class citizen, right? So you really need to pick your first-class citizens and bucket them together and then show them as groups in your navigation Now another thing that happens is we all have our backs in our enterprise solutions, right? We have role-based access control What I used to do initially I used to design a very generic Workflow and then I would say key is going to hear This role will not see this entry field that role will see these two additional fields Now what does that does it gives a very fragmented experience when they are actually using it when these roles are actually using it So later on I started restatigizing and for every role that I had I started redesigning that entire experience and What that helped me do is I could now efficiently, you know Customize it to that particular use case which made his job simpler. So always visualize human beings obviously we Read visuals more efficiently than text So whenever you have a very complicated environment complicated system try and use visualizations for the same Always assist. This is extremely important like Things like showing the right time zone. So if you're designing if you're giving an entry field for entering time We somehow forget to show the time zone of that particular object So I don't know if I'm entering my time zone. I'm entering the time in my time zone or somebody else's time zone So stuff like that like small instructions that make the Information extremely obvious is what I meant Now the the next point is efficiencies is is paradigm So usually whenever we are on a crossroads where you could design a simple solution But less efficient or you could design an efficient solution, which is slightly more complex always go for a solution That is slightly more complex But is more efficient because we have expert users because we have power users if you hide it Then the user will have to dig it deeper. So you're making the job even more difficult now So one crore to someone who guesses what that circle means Wild guesses. What are what yeah a glass? anyone else The moon Okay, so it's a car, but I oversimplified it So that's what I mean. You should never oversimplify stuff You should not make it so simple that it it it renders itself unusable choose trends wisely again like in gradients if if I Put gradients on the traffic signal. It's not going to make sense, right? So be extremely careful when you are choosing fancy graphs that don't mean anything or you're choosing gradients That probably confuse the user further be extremely careful with those kind of trends Automate the obvious if I'm making a list of everybody who's entering this building You should not let a guy sit and manually enter everything, right? You should just automate that entire process show your cards scan something and boom your information is in the system So always try to automate the obvious But also give an ability to create policies to the user So as a user I should be able to create my custom automizations I should be able to create policies that like if not then policies which help me You know automate things that the way I want them to always always give contextual actions Again, I think it's it's pretty evident what I mean like If you want somebody to feel a key You should provide an action that lets you go that helps you navigate to that page where you can get that authentication key and you know makes a job extremely simpler Always although personalization I think again, it's an obvious point on your dashboards or on your table I should be able to select what sort of columns I want to see in my table So always have those options. I should always be able to Select my widget see the kind of data they are displacing on my dashboard because every user is unique and every task is different, right? and This is this is another thing that I learned the hard way. I was obsessed with you know designing everything So when I was working with developers, there are some developers tool that are more efficient than The graphics user interface that we guys design But I somehow could not accept the fact that I kept on iterating on it kept on iterating on it And it all failed everything was less efficient than the developers to developer tools that they had So in the next part what I tried to do is I tried to design around it. So we tried to plug in Let's say CLI into the UI and we Started providing like support for that particular developer tool. So don't fight CLI is what I meant Mistakes are expensive. Obviously, you all know about it So show secondary implications like in our usual case if I say Clicking on yes will delete this file in enterprise. You have to say clicking on yes We'll delete this file, but we'll also render these four files unusable which will also render those next hundred files unusable So you need to show those impact to that depth Avoid the radical. So you all know the eBay story, right? So eBay had this yellow website They changed it overnight to white because Google was all white and it was trending Now they lost a lot of sales in that particular day. So they call this expert UX guy who found out why They got the yellow back because and people lost trust because the yellow was gone They thought the system is broken. The website is broken or something So that's why they lost the sale now What they did is they got the yellow back and they started reducing it to white One percent each day. So that's a gradual change and that didn't surprise people in the wrong ways the similar aspect you should probably follow an enterprise and Just because the enterprise shouldn't be boring You should use illustrations I mean, it doesn't need to look boring because illustrations also have a assistive quality They they tell the point across very swiftly They also act as visual triggers and they bring clarity to what that particular You know situation is about Try animation again. They have they act as visual cues. They have assisting Capabilities as well and at the same time they add delight Gamification again serious gamification like giving a list of things you need to do to you know keep your system healthy or Or like in mantra you have tears for Retailers like you have bronze you have silver you have gold So these kind of serious gamifications in a way they propel the user to you know Interact with the system in a better way and always break some rules So like design systems is what's happening right now. Everybody's working on their design systems Everybody's making their design systems very extensive But most of the time what happens is you can't have a single design system to design 20 different products, right? Nutanix has 20 different products. We have one design system doesn't work So whenever there comes a situation where the design system is not satisfying the need of the moment Feel free to break it feel free to make custom You know elements or components for the use case you have and provide delight So let's go back to the restaurant where we had biryani now you had your biryani You called for the bill the waiter gets you the bill, but with the bill he gets you a desert Now that's something now you were not even hungry, right? but you got you saw the cake and suddenly you're delighted you got something that you didn't know you wanted and That's what we should try in enterprise as well because whenever people tend to interact with enterprise. They have this preconceived Expectation that this is going to be boring my work is going to be boring So you should probably you know it could be like very quirky taglines It could be illustration could be animation could be anything But try and bring that unexpected delight and it will provide probably Make you make your customers more loyal. So yeah to summarize This is what we talked about today again. I did not put any screens intentionally I did not specify specific examples because I think we all have our unique use cases and Probably this might just help you in one of those and done And yeah, I mean I'm open for question That's my LinkedIn profile if you can scan it we it will be really nice to communicate with you guys on LinkedIn