Afroflare: Black in Tech

So we're going to get started and then when Noel joins we can just reintroduce him. So thank you so much for joining. My name is Maddy. I am part of the product content team as a technical writer. So that basically means that I write software documentation on the developer.docs website and I'll be your host for this. I'm part of the Afroflare ERG which is our in-fluid resource group for everyone from the African diaspora. And today we are joined by amazing professionals in AI and we'll be discussing the roles, how they use AI in the workplace and some professional insights. So the actual panel will be about 40-45 minutes and then we're going to leave the last 10 minutes for questions. So I'm going to start off with introducing each panel and the first one is GIFT. So GIFT is a senior developer advocate helping developers build fast scalable applications using Cloudflare's developer platform. With a background in front-end engineering and a passion for developer experience she shares practical insights on building full-stack applications, modern web tooling and life as a developer abroad. GIFT is also a LinkedIn learning instructor and international speaker known for making complex concepts approachable through talks, videos and writing. Thank you so much GIFT for joining this panel. The next one on my list is Anjola Adebowale. Anjola is a data and AI consultant at Microsoft specialising in the design and development of data platforms and generative AI powered applications with oversight of data engineering workloads. So she has contributed to large-scale programs such as the premier league as well as projects across both the private and public sectors with a particular focus on generative AI solutions. In addition to her enterprise work Anjola works with startups to help them scale their AI implementations, supporting early-stage organisations in building robust data strategies and embedding AI capabilities that accelerate growth and innovation. She has been recognised with the BTA rising star award and industry solutions EMEA star and rise awards given to teams who have built world -class solutions along with multiple nominations for her contributions to diversity and inclusion. Beyond her consulting role Anjola serves as co-lead of Microsoft UK's back employees group and Embrace University at which events lead, reflecting her commitment to fostering inclusive communities and nurturing future tech talent. Anjola we have a superstar here. Thank you so much for joining this panel. And then we have Ife. Ife is a prototyping architect, digital creator and builder at the intersection of technology, storytelling and culture. At AWS she helps organisations design and implement innovative AI ML solutions across industries and markets with a special focus on Sub-Saharan Africa. She is the founder of Coree, a creative club for young professionals pursuing both full -time careers and creative callings, offering community accountability and space for growth. Through her content and consulting Ife inspires others to embrace joy, creativity and impact while navigating multi-dimensional careers. Thank you so much Ife for being part of the panel. Beautiful. Then we have Noel. Noel has just joined. Hey Noel. Hey, sorry, wrong link. That's all right. Technical issues always happen but I'm happy that you are here now because it's just your time for your introduction. So Noel is a tech enthusiast and problem solver currently working as a technical account manager at NVIDIA. For nearly a decade Noel has had the opportunity to hold various technical roles in SAP. Noel is passionate about using his experience to boost tech adoption in African markets. He's in his early phase of his school so support and collaboration is appreciated. In his spare time Noel enjoys spending time indoors and in nature and likes to support Arsenal football team. Thank you so much Noel for being part of this panel. And last but not least we have Abdi Taima who is an advisory data scientist at IBM, leading generalized solution delivery across the UK and Ireland while managing sea level relationships and enterprise AI architecture. Previously he worked as a machine learning engineer at Lego in general, developing pricing models and as a data scientist at Capco. Abdi is pursuing his PhD in AI at the University of Birmingham and serves as lead AI instructor and academic advisor at Base Station. Wow, congratulations on your PhD. I hope that one is going well. And yeah one thing to brag about, so Abdi used to have a large gaming channel 10 years ago and yeah what happened to that one Abdi? 10 years ago I decided to pursue a career instead of just hoping that a life of YouTube gaming would make me money. Probably was the wrong decision now looking back but it happens. Okay cool, so we're going to start off with the panel and the first section of the panel is more career oriented and the second one is going to be more knowledge based. So this is a question open to all panelists. I would like to know how did you get in your common role and what do you enjoy most about your role? Anyone wants to go first? Sure, I'm happy to take this. So I studied mathematics and economics at Warwick and I was kind of tired of thinking so I wanted a break from numbers and anything analytical. I went into public relations and I was doing that for technology firms but I just found it wasn't really for me so I quit and I didn't really have any idea what I wanted to do after that. So it also coincided with the pandemic which kind of forced me into a career break but I applied for a role at Vodafone on their tech scheme and I was put into their software engineering grad scheme where I was for just under two years and then after that I applied to Amazon and that's where I am now. So it was, yeah, nothing was really planned. I just kind of tried to follow what I liked, what I didn't like and yeah ended up here. Cool and can I ask you a question about your education? Do you have any degree? Do I have a degree? Yeah, or do you have, did you, I think you said you mentioned that you went to Warwick? Yes, I studied mathematics and economics at Warwick but I didn't do any sort of coding degree, coding course or anything like that. I think I just kind of leveraged the transferable skills and I think my degree was a STEM degree so I think that was a signal enough that I could kind of learn and I think that's what it's been, mostly just kind of learning on the go. Okay, beautiful. That sounds, yeah, it sounds really interesting how people just are able to use their transferable skills to change careers and that's something that I always suggest because some often people think that they have to go back to uni and then redo the whole thing but sometimes you can just leverage your transferable skills. Anyone else wants to go next? Niall? Yeah, sure, happy to go next. Yeah, I got my role kind of bog standard like usual, just applied online. I guess for me the main driver was around, you know, again with a lot more roles being changed to be more AI focused. I saw my role doing the same, coming from, again, like SAP infrastructure, cloud infrastructure as a whole and I saw certain gaps where I felt like, you know, well certain gaps in my learning which I felt I could fill in that regard so I definitely applied for one of the, you know, sexiest companies when it comes to AI right now. Being a TAM, so I'm a technical account manager and at least, you know, it gives me a more general view of what's going on so I get to speak to a range of different customers, different workflows, different AI stacks, so it allows me to soak in a lot of that information and it still allows me to be technical as well. In terms of like education background wise, I did business computing which more or less is like computer science but without some of the difficult modules so you don't do maths, you don't do like algorithms per se. I ended up like learning it after the fact but yeah, it's pretty interesting. Oh, anyone wants to go next? I can go next. Go ahead, Gift. Yeah, hi everyone, so I studied computer science. I have like a traditional background and after school I delved into like front-end development so I was like front-end developer for a portion of my career and while I was a front-end developer, I really enjoyed sharing my knowledge whether it's through videos, blogs or creating content on YouTube and around three years ago, I made a video about Cloudflare Pages which is one of the products at Cloudflare and PM reached out to ask if I'll be interested in doing developer relations at Cloudflare. For me, I honestly had no idea about, I knew developer relations was a field but I haven't thought about pivoting to a different field so it was a good opportunity and I took it because it's generally focused on my passion for teaching so I was excited to join Cloudflare. So yeah, basically started as a front-end developer and now I work as a developer advocate. Beautiful, thank you so much. I kind of had a similar background as Gift in the sense I was a software engineer before and then I transitioned into technical writing so kind of cool. Abdi or Angela, who wants to go first? I don't mind going next. So yeah, the way I got into my role was I actually started up doing civil engineering for my bachelor's which has got nothing to do with technology, focused on soil mechanics, the fun world of that but then I decided to leave that behind and move into tech so I did a conversions master's course in computer science and that gave me the knowledge in typical stuff like data structures, algorithms and other stuff. And then I just applied for a job at an insurance company and I managed to join their actuarial data science team there and from there I learned all my skills really on the job. So everything that I've been doing over the last few years has really been because of that first job there and yeah along the way I've picked up some certifications from Microsoft, some of your data science and data engineering certifications and now I've decided to go back to uni to do my PhD so I think yeah that's been my journey into my current role. Beautiful, thank you so much Abdi and Angela next. Hi everyone, yeah I'll go last but so similar to Gift, I studied computer science so I did a placement year at Deloitte where I worked as a junior developer and from there I decided I wanted to work in big tech rather than I guess one of the big four and joined as a data and AI consultant and as an associate data and AI consultant and I essentially just worked up in that role so it was mainly just from my background in computer science and they typically were looking for like junior devs as well and then you also get to pick your specialty area so I decided to just pick data and AI over app dev or security. Yeah beautiful and I'm curious to know a bit more about your roles and if I want to start from you what does your what do you do in your day-to-day role as a prototype architect and this is a title that I've never really heard before I didn't even know it existed. Yeah sometimes I like the title more than the job but essentially it means that I work with different companies and we help them build rapid prototypes mainly using AI and machine learning and now obviously Gen AI technologies so as a prototype architect I'm responsible for enabling the customer engineers on our services but also building alongside with them and we do a co-build so that instead of just kind of developing a prototype and giving it to them so that it firstly accelerates their path to production but so that when I step off the prototype they still have the knowledge internally to actually move it into production so it's really exciting and yeah recently moved into working with customers based in sub -saharan Africa whereas before I was based I'm still based in London but I was based solely with UK customers and I think that context which has also been quite quite interesting and yeah a lot to learn. Okay that's super cool I yeah I think every that's the thing so a lot of people say that AI is going to replace a lot of people and jobs but I think it's actually creating more roles because I before if I didn't know that you could be a prototype architect using AI so yeah it's actually cool. Yeah the role itself I guess is an AI focus so the four areas that we work on are IoT, app development, data analytics and then AI machine learning is one of those four but obviously with the way the world is going yeah like 95% of my prototypes have AI machine learning in them these days. Fantastic thank you so much and I extend this question to all panelists yes I like I think I'd like to know more about your current day-to-day role and responsibilities. Who wants to go next? Gift? Yeah okay thank you so as a developer advocate my role the idea is to help developers building on the platform that I'm advocating for be successful and that involves me helping like creating technical content whether that's like videos tutorials blog posts so my day-to -day kind of changes because I context switch a lot sometimes I am helping the developer platform create videos for our YouTube channel sometimes I'm delivering a workshop at an in-person event or speaking at a conference the idea is to bring Cloudflare to the developers where they are whether they are in in-person events or they're online I also interface a lot with our community helping them you know use the platform and if of course they have questions directing them to people internally so for example the PMs or I sometimes also work closely with Madi as well she works on the CX team so if we have new content or examples that goes on the docs Madi helps as a reviewer so yeah the role's interesting enough for me because it stems from like my passion of teaching from teaching so I tend to do a lot of that components in my job. Okay beautiful and yes I have definitely reviewed some of Gift's pull requests in my day-to-day role it's good to work with you Gift. Now would you like to go next? Yeah sure so as a technical account manager I want to say I'm basically like a technical advocate for the customer so although there may be some use cases where we can step in and solve a specific issue or you know a customer might have a certain question like oh how do I convert you know this type of training job in AI to a different framework for example or something. If it's something that we could demo ourselves then you know I will go ahead and do it but we normally are partnered up with solution architects who may have like a deep focus in a certain area so for that we kind of manage the you know ongoing engagement with the customer whether it's like from beginning to end if it's a specific period of time or if it's like an ongoing engagement so we're normally aligned you know it could be based on products it could be based on you know industry but we're just there as you know someone if they've got any questions or it could even be like if they've got you know support for example and to some degree sometimes they may have like a support plan or something so we also liaise with support teams to ensure that there's an escalation point and if we need to bring you know like deeper engineering into the conversation or management then that's something that they can lean on us to do so we're kind of like a mixed bag of different roles and again you have some TAMs which are more generalists and you have some that like focus on a specific area so in my case it's mainly always been around I want to say like you know cloud infrastructure and yeah and now more or less like AI infrastructure so the various components. Beautiful, thank you so much Noe. Abdi or Angelo, you want to go next? I'll go next so my day-to-day is split between more being a sales engineer so working with some of my customers to help them embed or scale up some of the IBM softwares that we have in the AI space but another part of my day is doing the AI research that I'm doing and that involves a lot of reading so finding what papers there are that have come out in the space that I'm looking at and then asking AI to summarize it and tell me if it's worth reading and then reading that if it's interesting but yeah that's that's my day -to-day. In terms of my day-to-day I work closely with customers so essentially the consulting arm within Microsoft so we just build solutions for customers anything from data platforms to anything generative AI so my day-to-day is essentially that either we're doing pre-sales trying to I guess pitch the idea to the customer or build out and design the solution or POC of the solution or if it's like a larger scale program you'd be looking at building or co-engineering it with their team so definitely a lot of development work but also designing and also just thinking I guess long-term strategy on what the customer wants. Beautiful, thank you so much Angela. So now we're going to move to the more interesting section of the panel and we're going to discuss more about AI and productivity. So earlier this year I was at the London Tech Week and this gentleman was explaining about like how many startups, small companies are selling softwares saying that it's just AI but in reality they're selling AI software but in reality they're actually selling a software that just automate tasks so this gentleman was emphasizing about the difference between AI and automation. So now I'm going to write this question to Abdi. Can you explain like on five what is AI and how do LLMs work? Sure, I'll try to be brief and explain what AI is and then I'll move into explaining a bit more about LLMs or models. So think of AI as just a way for machines to learn from examples that we've given and to act like us, like people. Now there's different forms to AI, there's the stuff with chat GPT and generative AI that's exploded over the last couple years but there's also the wider world of deep learning and all the other stuff that comes with that and then all the other things that have been around since the 1980s including symbolic AI and a few other things. So it really is a big discipline and a lot of people would even argue that it is just computational stats so it's just statistics with some compute with our data put together. Now if we look more into what LLMs are, so think of when I say LLMs you know chat GPT or Claude or whatever you might use and they've come around by we've just been teaching these computers to learn from all the world's information so things like Reddit, from books online, our tweets and things on X and from that we've been able to simulate what a brain is because our brains and how LLMs work, they work in a similar way where they're trying to create summaries upon summaries and just abstract knowledge and then give us an outcome one word at a time. So that's really how it all works. I'll stop there because I could go on for days but it's very interesting technology. Yes, thank you so much and I agree with what you said and I think the main difference between AI and automation is that automation is missing the learning component so it's just based on predefined rules whilst AI is actually constantly learning by feeding data. So just wanted to clarify the difference. I want to discuss about productivity because a lot of companies right now are encouraging employees to embrace AI as part of their roles to make them more productive, to make sure that we actually spend our time on the most critical activities. So I'm going to direct this question to Ife. How have you used AI to boost your productivity at work? I think in two ways. I think the first is more productive, the second I'm actually making me less productive because I spend time just going over emails that I wouldn't usually, I would just send it. But my favourite tool is called Amazon QCLI and it's basically a terminal or IDE where you can look at your project structure for example, you can ask it to review your code, you can ask it to suggest code for a new project. I feel like the applications are really endless and I think being a prototype architect and I think as Angela was explaining what she does, I think there's probably a lot of crossover there. You have to oversee the whole project so this can be from qualifying it to suggesting what the architecture is going to be to actually building it to writing a prototype report afterwards which basically details everything that was done in the prototype and then what the suggestions are after. I think the core part of my role that isn't really replaceable by AI is the middle bit. When it comes to qualifying it, writing the scoping document, writing the prototype report, I don't really think that I need to be spending much time on that so I think that I use QCLI to do the bits that or to at least assist with the bits that I feel like are basically repeatable from prototype to prototype so that I can focus on the actual build. That would be my recommendation and I think if someone wants to get into coding as well, I would also use it to just suggest starter projects that you can learn Python on or things like that. I think it's a great tool. Beautiful. I actually had no idea that Amazon had this tool so great to know. Thank you so much, Ife. Gif, did you want to add something? Yeah, I'll just add I use Gemini a lot. I also use Winsoft so I create a lot of demos for my work whether that's trying to create a demo for a tutorial or for a video and I use Winsoft. Winsoft is like a code editor that has different LLMs or models in it. For example, we have Cloud Sonnet or Gemini that you can use when you're writing code. You can access to help you with your code like review your code for example or help you also add functionalities in your code. I use that as well as Cloud Code. Cloud Code is a CLI tool that does similar things with Winsoft. These are some of the AI tools that I use for my PhD and it really helps boost my productivity. Thank you. I have been using Gemini as well lately. One example I can mention where AI has saved me so much time was I saw then a user request. This request was to add a table where you just can compare different types of deployments. What I did, I asked Gemini, can you please compare page ABC and create a feature table based on the content that you find on each page. Gemini as a result created a CSV file that you can export on Google Sheet. Obviously, it wasn't perfect because you still have to tweak it but it has saved me probably a two weeks of conversations with stakeholders. Now, instead of me having to go to stakeholders and ask, okay, can we do this together? I can now go to stakeholders and say, okay, this is what I have. Can you please review it? That saved me basically two weeks of conversations. Within 24 hours, I managed to publish the comparison table and that has solved a customer problem within maybe a week or even less. It's super helpful. I wanted to ask about how can someone learn more about AI? Noelle, what do you think? Yeah, definitely. I think, first of all, I think Google and YouTube is always a good place, myself anyway. I think, definitely, I would say reach out to anyone in your network who's already in a role, maybe a more AI-focused role. From an independent perspective, I think starting off with understanding what AI is and then also the step. You can find a lot of information online, I think, like free certifications. For example, NVIDIA's got a basic AI course from my basics to Gen AI and I can plug it in the chat at some point. Also, Google, Amazon, Azure, a lot of companies have their own kind of 101 of AI. One thing I would champion is actually use it. Content is one thing, watching things, reading articles, but actually giving it a go yourself. Whether it's from a simple LLM perspective, so using chatGBT or perplexity or something initially, but then also applying that in your real-life scenario. Just think of something you do on a day-to-day basis that you think could be either accelerated, so maybe it's taking you an hour to do, think about how you can use maybe an LLM to reduce that, and actually ask it, so prompt it, like, how can I make this more efficient? It might not actually solve it for you, it might give you pointers or an answer. So that's my brief answer. Yeah, definitely. I pretty much agree with everything. I think the best way to learn more about AI is to actually use it. Anyone wants to chime in? Okay. I'm just going to plug a resource. It's called AI Avenue. I'll post a link in chat where I think the chat is disabled, so I'll just share it in Q&A later. It's essentially a video series coupled with tutorials that shows you the basics of AI. It teaches you, like, your five, what AI essentially is, and I find that for people, specifically non-technical people that are, you know, trying to understand what the whole AI trend is, this is a very good resource. Let me see if I can just post it here in chat. I think, if possible, Maddy, I'm not sure if you can post more than the host and the panelists. You might have to open up the channel. Maybe after the session. Maybe after, but the URL is AIAvenue.show, in case anyone is interested in checking it out. But yeah, it's a good resource to learn about AI. Thank you. And the one thing I'll just add to all the amazing insights is just using the AI tools to summarize what the latest is. There's always new things coming out, and even, you know, I'm a researcher in this space. I find it difficult to keep up to date with everything. So, I just tell the AI, hey, go find out the last three weeks what has been put out there and summarize it and let me know in easy language, and I find that to be very useful as well. Thank you so much, all. I wanted to ask a questions around using these chatbots, like ChatGPT and Gemini. So, what should people consider when using these tools? And I ask this question because a lot of people have concerns, like they don't really want to use these tools because they're concerned about their data. So, their personal data being exploited. So, do you have any suggestion on how to, on best practices on how to use ChatGPT, Gemini, and other tools like this? Sure, I can jump in here. So, I think like most obvious is obviously don't put anything that's like identifiable in there, like even addresses. And I just try to, especially with working with customers as well, like I'll just redact their names. So, you just put customer instead of the actual customer itself, or like name instead of the person's names. But also, I think on a personal level, I think ChatGPT can do a really good job at like, I think, validating you. And like, even if you just have issues with friends, like it's going to be like, yeah, you're so right, like they don't deserve you, all of these kind of things. And I think you have to be really careful about like replacing ChatGPT with like therapy as well. Firstly, just from like a human perspective, but also because like that kind of, that information can get summoned to like legal cases and things like that. So, if you, I don't know, you'd like killed someone, or you went to ChatGPT and was like, what should I do? Like that can be used as evidence. That was a very drastic example, but it's just the one that came to my mind. So, I think, yeah, just anything that can point back to you in like a negative way, I would avoid putting into any AI system. I definitely agree. Noelle? Yeah, I was going to say, just to add on top of that as well, like the reason why you shouldn't put personal data in, especially when it comes to public LLMs is because they're training against the data that you put in. So, somewhere down the line, you know, that could impact yourself. So, that's what I wanted to add. I think just building on top of that, also just make a habit of turning it off in terms of allowing your data to be used in terms of training, and also make sure that you're using it to support your decision-making. So, if you are using LLMs, you are trying to use it to get more information or to support you in things. It's just supposed to help you make, let's say, decision, help you make the decision, but it's not human judgment. So, you can't really replace that. So, just make sure that you're thinking about things like that, and also providing context when you are working with LLMs because you are seeing more cases of things like hallucinations. So, just kind of making sure that whatever you are asking it and doing, keep within a certain sphere. So, even if it's adding more context, some structure, some examples in terms of what you want, just to make sure the output is as close to accurate as possible, just because obviously people are using it for different things, just to make sure you're being safe as well, things like that. Yeah, beautiful. And a different question, has any of you built any exciting projects using AI, ideally in the workplace? Oh, go ahead, Nil. Yeah. Oh, okay. Yes. So, I think the most exciting one I've done this year was probably like an automated insights engine for a bank based in South Africa. So, they wanted a way to kind of automate the collection of insights from their competitors that they could use to advise their marketing. So, that looked at kind of scraping, and this was all public data, so scraping any new products that they had released. I think what was most exciting was the browser automation element, which is essentially getting AI to act like a computer on your behalf. So, instead of what the marketing consultant was doing at the time, going through Twitter herself, like taking screenshots of interesting tweets, interesting pictures, we instead prompted AI to browse Twitter, or like browse YouTube, and like basically gather those insights. And I just, yeah, I found it quite exciting. I think that that use case does have a lot of applications as well. I think, interesting that it came from the idea of privacy, because you have to be very careful in terms of like competition, and what is ethical and what's not. But I think that what we were looking at was just public data, and it was just information that was actually already available, and it was just doing that like in a faster and more dynamic way. Basically, that sounds really exciting. Anyone else has built any interesting project using AI? Yeah, I can share. So, we've been speaking a lot about LLMs. I feel like the next big thing in that area is agents. So, AI agents. A few weeks ago, I built an AI agent for my own personal need. So, it's a travel AI agent. And what it does, it essentially helps me, so if I'm going on a trip, helps me find like flights, helps me find hotel in the location. So, all of this, I would typically do myself. I'll typically research like hotels or flights, to a specific place that I'm traveling to. But I built an AI agent using Cloudflare Agent SDK, and this essentially helps me with just a simple prompt, plan my trip. And it's also implemented with obviously booking.com for booking the hotel and flights API. So, if I decide I'm going to go ahead and book that trip, I also have that implemented in that the agent handles everything from flights to hotel booking and also recommends things that I could do in the city that I'm traveling to. So, this is just a personal project that I was looking to explore. And I used this as a demo in one of my recent presentations. Yeah. Beautiful. And so, if you mentioned Cloudflare, I wanted to talk about this new tool that Cloudflare announced on July the 1st, which is called Papercrawl. So, Papercrawl basically allows, it's for content creators, publishers, writers, and allows them to get paid to have the content script. So, basically, if you want an AI crawler to have access to your content, then you can ask them to pay you instead of getting your content for free. And I want to mention this because this is one of the many things that we do at Cloudflare to make the Internet fairer. I'm definitely going to share the link just before the end of the panel. But if you're a content creator, you may want to check this one out. Okay. So, I'm just going to ask one more question, then we're going to go on to the Q&A. So, I'm going to open this question to all panelists. So, what advice would you give to professionals looking into a career in AI? I can take this. I would say definitely keep an open mind because I think building in the AI space, it moves very quickly. And there's always new advancements in the space. And especially if you're building for customers, they might expect you to already know. Even if it came out on a Monday, they want you to know on a Tuesday how they can deploy it. So, we just keep an open mind in terms of ambiguity and learning. Because I do think a lot of roles and teams are looking for people that are willing to learn rather than having people extremely skilled within the AI space. Because there are a lot of areas that people are still learning and still growing within the space. So, definitely keep an open mind. And also, I guess, focus on making sure you're the kind of person that they can just put you in a room, give you a set of tools, and you will try and figure it out. I think that would definitely be a key one at this point in time. Yeah, just to add on that as well, I want to say that AI is like pretty much any other domain within tech that has been hot at any given moment. So, also be open to the different parts of stack when it comes to AI and the support in components. For example, I don't necessarily, or I'm not someone who works on the AI workload level per se. I can run some jobs. I can run some examples. But my role is heavily rooted in the infrastructure. So, if you think about the models that are behind some of these AI tools, they need proper fast networking, for example, for it to run. So, there's a need for networking people in AI. There's a need for developers in AI. There's a need for storage people in AI. So, anyone on the call who has a specific domain within tech already, just think about your area and how that can translate into the new phase of things. I think that's quite important. Yeah, I think that's a great answer. I was just going to, I'll just add on that. I think aside from even in AI, it's important to just think about what you're interested in, in general. Like what area of, I don't know, what kind of customers do you want to work for? Is it retail? Is it general? What kind of customer problems and things like that are you interested in solving? Because those are the things that are always going to be there. Like people are always going to have problems that need solutions. And right now, everyone's trying to use AI to answer those problems and to build those solutions. But it's got to be something else soon. So, I would say to kind of work backwards from the area that you're interested in. And as Noel said, figure out how AI can help those solutions. And then also think about where AI is going. So, I think we mentioned data being so important to training AI models, to a lot of the AI solutions. So, that's something that's always got to be important. So, understanding how data works and how it influences the rest of the ecosystem. I think to summarize, focusing on what you're interested in, in general, and then looking at where AI is going in that area. And that would be my advice. Thank you. So, I'm just conscious of time. So, I'm just going to go straight to the Q&A. We have some really good questions. So, someone is asking, what are good ways to evaluate different AIs for different use cases to know which one to use when? When exploring outputs from different LLMs, especially for PROM tasks that produce long or complex responses, it's sometimes not feasible fully read and analyze multiple responses from different AIs at the start. Anyone wants to take this one? I'll have a go. So, I'll give two answers. The first one is the, what I would say is, it's my day job is to help people with that. So, we have some technology in IBM that helps solve this problem. But if I take that hat off and I put the IBM, my normal AI researcher hat on, I would say that there's a lot of open source evals available online. And I would try to say, look at what recent research there is in the type of use case you're looking at, and then just use that, reimplement it. There's a ton of open source evals you can pick up and use. And if that is too difficult, just use the LLM as a judge approach and just fire your problem into another bigger model and see what it says for your evals. But yeah, that's what I would say. And of course, if you want to see IBM's answer, just ping me offline and I can give you a long talk about our product. Thank you so much, Abdi. A question from Bradley, what opportunities do you see for AI to advance equity and what risks concern you most? And I want to take this one. Sorry, could you repeat that question? Yeah, sure. So the question is, what opportunities do you see for AI to advance equity and what risks concern you most? Okay, I can take that. So I do work across some different African countries, and you do see the idea that AI does empower the average man. It's allowing people to be able to do more, achieve more, anything from even starting a business. So there is definitely more AI can offer to different people across the world, regardless of your background, except obviously factoring, except you factor in, let's say, a digital divide. But if you have access to a phone, computer, it's going to help in terms of that. But one thing for me, especially even being a Black person, is the risk of we all have to be in the space, and you have to make sure you're actually actively building in the space to make sure that it factors in people like us, that additional context. The bias isn't going to be there. The hope is that the bias shouldn't be there a couple years from now, that when they're building out all these LLMs, when that's happening, are we in those rooms? And are we actually factored in? Because that will become a problem over time, and it can easily become like a long-term socio-economical factor that could affect people. So I say that's one of my major risks in terms of making sure we're in the room, and also making sure that we're a part, and we are being added into whatever's being used to train the LLMs each season. Beautiful, thank you so much. So we have time for maybe one or two questions. Could you point us to the top certification courses you've seen done so far? Anyone has taken any certification courses? I haven't personally, but I would probably look at Google, since they basically created Gemini, and also Coursera. Usually when it comes to tech courses, I definitely suggest Google certifications and Coursera, because they tend to be really good. Yeah, just would add AWS there, obviously. But yeah, no, I'm serious. The AI practitioner, I think, is quite a good entry-level certification to go for. So I think it's, especially if you've done some work with AI, it's quite a quick win, so I'd recommend that. And if you wanted something a bit more deeper, then the machine learning specialty. Beautiful, and we have time for one question, a quick answer. We have two minutes. So with so many new AI tools coming out these days, how do you stay informed on your AI tools without getting overwhelmed? Or how do you figure out which ones are actually useful instead of just hype? I can take this, actually. I think you should just look at what you do on a day -to-day. So understand your own personal needs, and if you're in a team, what are your workflows, and then work from there. As long as you know what the output is that you need, I guess the speed you want it, when we work with customers, we always start from that. Create the use cases in your own personal life as well. So if you're coding on a day-to-day, like I use LLMs just for coding. So I know the specific ones that would work best at the speed I want, accuracy I want. And then I focus more on building out agents that would basically do that specific task that I want, instead of jumping from one to another. Because I already know, okay, I have this specific agent, let's say within Copilot or within GPT, that does exactly what I want. And you just keep going back to that specific agent. You can do that just by creating no code, creating an agent, give it exactly the instructions every time you paste that data in. So just pick the key ones that work best for you. So I know some people would say, GPT is really good for coding, but it's really good for humanizing information. So I guess, what are your key pain points? And then it's very easy to kind of select the one that just works best for you with that problem. Beautiful. Thank you so much, Angela. So we are at time, it's 5 p.m. I'm going to wrap up this panel. So I wanted to give a special thanks to the team for helping me put this event together. And also Tana, who has helped me with the designs. And also the team for the technical support. So we have Mio here. Thank you so much again. And yeah, see you soon. And reach out to all the panelists on LinkedIn if you want to connect, if you have any more questions. Bye.