How Swisscom Transforms and Innovates with AWS and AI | AWS Events
good morning morning Mark Manisha Roma welcome to the AWS studio i'm Roben Merz principal solutions architect at AWS and today we're going to talk about AWS transformation mark please yes so I'm Mark Mark Dner uh I'm the CTO of Swisscom so I'm Manisha Day i'm the product owner at S S in Swisscom and I'm particularly interested in really driving innovation and efficiency and therefore very passionate about generative AI what is S sur so at Swisscom we practice uh Swisscom reliability engineering which you can map to definitely different aspects so you have network you have software you have system and um that's exactly what we do here to focus on the reliability and the stability of our existing systems i'm Roma i'm leading our uh transformation efforts uh to change the way we operate our services sre is one part of of it but there are many other aspects so trying really to rethink how we fundamentally operate uh on the daily basis our services taking inspiration from as many uh good initiative as many good technology geni is one of them so taking all that and leaning us forward all right thank you so um Mark you've been here uh last year we had a fantastic uh conversation on transformation and especially in networks and infrastructure now I realized this morning that when we started working together Swisscom and um generative AI was not on the radar and suddenly it popped up and it's it's it's definitely it's changing a lot um what is your what is your view on on on on on data and and and AI and on on that technology how does it how has it changed Swisscom and and how do you see it um shaping the company for the future i think it did change a lot right and it is continuing to change just started i would even say um one example is um when we started um the journey with you one of our goals was to uh create our our uh what we called one data platform in the cloud obviously um and that was a technology decision it was a technology project um with the um with Genai uh creating impact all of a sudden business is asking why are you not faster mhm um another way is the perspective on speed on innovation jas showed us how fast innovation is happening i mean the the large language models are improving themselves by four by a factor of four every half a year and therefore uh innovative and um as Swisscom we strive for that means to be adaptable at a much higher speed than in the past and that is due to um how can I learn faster how can I build an architecture that can adapt faster and how can we organize ourselves better and the last one is maybe even reflecting on the first one is uh getting us even closer together so uh we already did work quite well jointly between our organizations so technology and and for example B2C um but with Gen AI um we need to scale we and uh having a team in technology doing AI was not about scale having a team in technology having a team in B2C this is scale so the way of working is changing because of that yeah I find it very interesting how when when technology changes and you need to adapt this really brings people closer and this really test relationship and this is where you see this is how you build high performing teams uh speaking of high performing teams that's very good um Manisha Roma both of you have been uh have decided to jump heads on into uh technology and into AI and genative AI and how it's changing what you do um Manisha I'm going to start with you can you tell us about uh about your projects what's your driving so I think um the first important factor here we are in a transformation phase exactly as Mark mentioned we are trying to innovate we are trying to adapt ourselves to the ever evolving technology and the best thing that you can get out of this transformation phase are really some concrete tangible results which enables engineers to see or realize the potential of generative AI in the world of telecommunication and this is what exactly we are headed towards so we have an uh initiative which is called the SR copilot quite uh obvious because I come from the S sur world so what this um assistant is about to do is going to speed up our transformation journey from transforming from a telco to a techco which means that this would enable engineers in their own s transformation journey because now in Swisscom we have a lot of s knowhow in the form of documentation in the form of really extensive skill up trainings and we can use all that to feed that LLM so that it is now responding in a tailored way to every engineer to enable themselves to know How would I build my own uh uh application which platform do I use so consolidating and standardize standardizing practices coming to tooling coming to coding as well because we have a lot of teams at Swisscom which handles problems the same problem in different manner and we want to standardize all that practices we want to bring things together and we can do that by reusing sharing the code inventories not by reinventing the wheel from scratch and the SR assistant or the co-pilot is exactly aiming to do that interesting aspect of the SR copilot why we started with it is also on a psychological aspect because you mentioned people and culture which which is the most important aspect here and here we wanted to speed up the transformation journey and we felt that people are so comfortable in talking about their own transformation goals and how they want to get where with an LLM rather than a human being so why not use that power and try to speed up the transformation so um that's one of the initiatives the second one which I would talk about I'm very happy that I got to interact with a couple of more telco operators here at MWC and we are all more or less headed in the same direction so I'm uh looking forward to having synergies there so we are developing a um RAN network assistant which will be able to understand the network health in terms of the radio network so the idea now is to really prototype a solution so we have already prototyped the solution it's small we want to iterate over it at the moment it is enabling engineers to get a holistic view of their network health by uh getting data from multiple data sources doing the correlation between the KPI metrics so in the end we are simplifying the lives of engineers who don't need to now open five different tools to form a correlation and ask okay how is my network in I don't know some place in Switzerland performing so it can be done through the LM as a next step we want to take it forward we want to transform the reactive nature of the LM to becoming more proactive and understanding what it outputs to the engineer what is the network health looking like so I can suggest measures to the engineer already beforehand so I think um these are the two main projects which I've been working i I I like I mean this you came up with something new you came up with aspect of personalization of a shared knowledge i I find it um I mean this is something we talk in in marketing topics but not in that that's super cool so I I look forward all of us working on getting that project to production roma um next to you tell us about tell us about what you're driving so we have started a quite large project on using jai to change the way we operate our services as I was saying before so think about ITSM kind of concept uh change management incident management that's that's kind of the scope so ITM it's more than those two obviously and we are really not checking all and we have started to build use case on to transform based on that and it's bringing a lot of new opportunities uh at first we were last year more focusing on Morag approach simply chatbot uh conversing with the documents bringing value directly to the people and and the and as the time is going what's kind of amazing is back to what Mark was saying or fast the technology evolve now we are speaking about agent agentic concept I don't know yet to tomorrow or what we will discuss about but the speed of innovation is really crazy and it makes it super interesting so now we are really focusing on uh agentic approach so having not only the the genai answering questions but really performing actions in behalf of our engineers the goal being our engineers should spend far less time working on the processes documenting what they do uh analyzing risk uh mitigating or thinking about possible mitigation measure they should really have as much time as possible to bring new features to bring new values to our customers and everything that we can then offload to Genai along the operation chain is basically a huge win first because of the time but also because of the increased quality that we get we are quite good with those processes but it's it's quite difficult for the team to every day redo the same kind of manual work to to fulfill quite many uh compliance aspect we may have assurance aspect we may have and jai here I believe would radically change the way we we approach those those things it will be done for us for our engineers in an extremely high quality way maybe long term we will even I don't know not need anymore to access so many tools as we have today you could probably interact with the rest of the company a lot more of our genai agentic systems so it kind of it may abstract quite a lot of the the work for our engineers so the focus here is really changing the way the engineers operate so that they have more time for added value tasks that is quite these are quite massive changes how do you bring people along this what has been your strategy over the recent because we we started talking about these things maybe 18 months ago how have you driven people what have you what have you learned over the time to really get people to work on this technology manisha would you like to to to share so I I I think as I mentioned there are a few learnings that I that we can take out of this journey now and one major uh thing is that not to really run behind the um uh let's say the highest compute that is available to you or the biggest models that you can but it's more important to really understand what my people in my organization need what would benefit them so how can I really integrate AI workflows into the legacy system into the exist existing uh work scenarios and how would it really make them them as in mean people in my organization produce more impact do what they do today but with greater efficiency so I think that's what we have been doing back at Swisscom so it's not really about which models we use but it's more rather understanding people their needs also understanding our systems are our systems really ready for data with does that have that data readiness to have the generative VI solutions or not and then really driving this together with people co-sharing collaborating and um taking these kind of initiatives to greater heights exactly with strategic partners like AWS all right thank you Roma Manisha I already mentioned a lot of super important aspect what I can maybe add is the question you ask is extremely valid so people and culture are key in those transformation project the first things I would like to state is that's key but the technology is also important and Working with a solid technology partner helps a lot because you can stop thinking a bit about the tech stack and how well it works and really focus on those transformation aspect so that's that's a prerequisite you just need you need to approach for me the people and culture aspect from different angles you obviously need a strong leadership support to that transformation we obviously have that with Mark that's key but this is also true along the world chain of leaders that we have that we we need everyone be able to see the potential and embrace the transformation you also need a more bottom up approach you need some engineers in every team that have time to investigate try fail try again succeed so you you need to bring both those direction at the same times and what we do to support that is also to create trainings material we we create events we have different communities where people meet exchange discuss together so this discussion across let's say the horizontal teams that we have the horizontal levels different teams from different um different units that we have but also the more the the discussion between engineers leader you need all those discussion to happen and what we have to offer is a platform for those discussions to happen and create through those discussions the motivation the create a common vision on on those aspect mark I I I I want to ask this so what's your secret source on because at the end of the day you have to be efficient you have to deliver features products make the network run but what I also hear is that you need to create an environment that provides people for this for these places where they feel safe and and they can they can go and experiment so I mean how how do you do this because I don't think it's self-evident we have a great company culture that that allows and enables that in the first place um and that's been like that for decades now but secondly um new technologies or change you can't convince people the only thing you can do is um get them passionate inspire them mhm and uh and that is is is a team effort so it's not me inspiring that would would not work right i mean the team is inspired by their teammates by people using it and saying it's great uh you need to try I have something for you um and then once once this starts to grow then at a certain point in time you you can build a structure around that and make it a force um but in the beginning is always inspiration and passion okay and that's what I love about our company it's really passionate we are passionate about technology we were passionate about our customers and this is driving um the teams right and and how do you recognize that moment where you see that for maybe a specific topic you come to that point where where you have this sort of like critical mass and and you can really like do you have to do something or can you just nudge it into the right direction the risk or let's say the the challenge at some point is um and and Jennai is a very good example everyone got passionate I mean We love technology we love to be at the forefront all of a sudden we had hundreds of projects running which is not efficient duplicating things um and risky because I mean you're dealing with data um and so at a certain point we we we understood okay um we've got enough passion we've got enough inspiration that's not the challenge anymore now we need to structure it to make it force and this is what we did the last half year so created an AI strategy um a data strategy um a data and AI target operating model how do we want to work to make that to channel that passion channel that force um for the greater good can you share with us how how we've been helping you i mean we we provide services we we provide people we work closely with your team but can you highlight on these options what have been what have been the things that that have been truly helping you to move forwards with those projects there are quite many aspects so obviously you know your services super well and you are you know how to build or to use them in those kind of concept extremely well but something that really helped me a lot in the last 6 months is all the other aspect that you bring what others are doing which kind of alternative technology are there because you do bring this overview of not only what you have as a service but also what's happening in the world ecosystem worldwide and this is helping me a lot to think about the the options that we have what what is the best way to serve the purposes that we want to um uh to to achieve and and then that's for me amazing not only the the part that you deliver in your services but the knowledge you bring on the wall gen ecosystem this is this has helped me a lot lately okay I guess that's a testament to the people working uh working with us yeah I think I I agree with Roma here completely and I would just like to add probably a few points here so um you know the services pretty well you're the experts here so that's given but What I value the most is that you guys are there from right the very from the very start phase of brainstorming the solutions and how it applies to Swisscom context i think that also um makes a lot of difference that's a game changer because yes there are a lot of generic stuff but not everything would produce value for Swisscom would not be driven or would not drive business for Swisscom and I think to tailor that generic stuff how we can tailor it to Swisscom needs I think that's been really amazing and then of course um um discussing with you till the point of which types of feature capabilities do we need on AWS right now which kind of solutions we are building what would be beneficial to us I think that's um been an amazing support with you guys i'm looking forward to continue it okay thank you i mean I'm a very strong believer that that doing this transformation this is about technology and that's about people and I think your your your your statement here are really a a very very bright uh testament to that um as we're as we're closing Mark any any any last word something you want to share with the others out there that are in those transformation starting on transformation maybe you have to start from scratch tomorrow where would you get started or do different I think I I never look back and say what would I do different because um yes you need to learn uh in that context but I think um the key theme of Genai the um coming coming alive is the incredible speed that all of a sudden happened if you think okay that's overwhelming you can't um then then look at what people do how many people are all of a sudden becoming experts in a very short term and I mean becoming an expert in the past in whatever domain took years to master that and uh now becoming an expert in at least applying genai is much faster in itself and then the other part and um that is and was inspiring for me for example in working together with you is um you can do that in any organization sometimes we think uh only small organizations can be fast but when open AI launched you were taken by surprise as any anyone else right you didn't have your lineup to counter that a year later we went to reinvent and it was a firework of genai and uh seeing a huge organization like yours moving that fast was an inspiration to say okay we can do that as well we have to and that's the passion part we want to and I guess so if you're afraid of moving just do it's fun it really is and you you you will find a lot of people dragging along like like we do we're super uh excited that we can share that transformation and this uh and and and and this technology journey with you um I'm super looking forward to the next 12 months everything we're going to achieve and I'm super looking forward that we'll be here one year from now um talking about what we've learned and what we've produced we here live in Barcelona i'm Ruben from Adess and see you next year
2025-04-28 13:45
