Even the best AI use case is useless if your people cannot build it
At Wells Fargo, 75 percent of searches now go through an AI agent, cutting search time from 10 minutes to 30 seconds. Yet only 24 percent of leaders say their companies have deployed AI organization-wide, and some tools boost output while quietly eroding skills. Jessica Perez, a Microsoft Technical Specialist with over 18 years in AI and cloud architecture, asks which AI ideas deserve building, and which build capability instead of dependency.
See how the AI use case radar scores ideas on business value and ease of implementation
Jessica Perez and her Microsoft colleague Ingrid Aaseng explain how organizations can discover, ideate and prioritize AI use cases, from spotting problem signals and adopting AI-first thinking to scoring ideas on a use case radar. They argue that people, culture and AI fluency decide whether AI creates value, outline a phased path toward human-led agent teams, and close with a low-code Copilot Studio demo.
“But even the best use case is useless if your people cannot build it, use it or govern it.”
“With AI first thinking you start with capabilities, not problems.”
What you will learn
Spot the signals of processes ripe for AI, from repetitive rule-based tasks to knowledge trapped in documents, emails or experts' heads.
Apply AI-first thinking by starting with capabilities, assuming abundance and designing systems rather than single tasks.
Evaluate whether an AI use case acts as steroids, sneakers or a coach for human capability.
Prioritize AI ideas with a radar that weighs business value against ease of implementation.
Plan a phased path from a first low-code digital assistant to human-led teams of agents.
Key takeaways from the speaker
Discovery & Ideation
AI-First Thinking
Use Case Assessment & Prioritization Tools
Agentic AI and the Future of AI
Details
Level
Beginner
Format
Talk
Technical depth
Introductory technical
For
Executive, People manager
Length
34:00
Published
July 2025
Questions this talk answers
Where should organizations look for AI use cases?
Look for signals such as repetitive rule-based tasks, data-rich but insight-poor areas, processes relying on expert judgment, painful customer journey moments and information trapped in documents, emails or experts' heads. Alternatively, flip the script and start from what AI can now do.
What is AI-first thinking?
It starts with AI capabilities rather than problems, assumes abundance rather than scarcity, and designs AI-augmented systems instead of automating single tasks, for example managing the whole vendor relationship lifecycle rather than only invoice processing.
How can you tell whether an AI solution builds or erodes human capability?
Use the steroids, sneakers and coach model from Microsoft Research: steroids boost performance but erode skills, sneakers save time without changing you, and a coach helps you grow, which is the gold standard.
How do you decide which AI use cases are worth pursuing?
Map each idea on the AI use case radar by business value (customer need, differentiation, revenue or cost, strategic fit) and ease of implementation (data quality, privacy and legal hurdles, integration complexity, skills). Then document strong candidates in a business use case template.
How can someone start experimenting with AI agents today?
Pilot a high-impact, low-risk use case, for example by building a digital assistant in Copilot Studio without code, giving it a description, instructions and starter prompts so that short requests produce useful output.
About this talk
AI is no longer a future bet—it’s a present imperative. This presentation will explore how AI can be leveraged as a strategic differentiator, transforming traditional task automation into agentic AI systems that proactively identify opportunities, initiate actions, and collaborate with humans. Learn how to discover high-impact AI opportunities, generate innovative use cases, and prioritize them for maximum business value.
Don't miss this opportunity to gain insights on building an AI-forward culture and scaling AI as a business capability. This talk will provide you with the tools and strategies needed to stay ahead in the AI-driven landscape.
AI-first thinking: start with capabilities, not problems (89 s)
Is your AI steroids, sneakers or a coach? (90 s)
The AI use case radar: value versus ease (60 s)
And on the topic of discovery and ideation of AI use cases. My name is Ingrid Åseng. I'm a cloud advisor and Azure specialist at Microsoft. Can you shift the screen, Jessica, to the next slide? Yes. And I spend my days advising large organizations on digital transformation and innovation in the cloud. I have a background in management engineering and software development. And with me today, I have my beloved colleague, Jessica Perez, who is the technical specialist at Microsoft.
She has her key competence within product development, artificial intelligence and cloud architectures. She has more than 18 years of experience across geographies such as Latin America, Western Europe and the Nordics. And we have a fully packed agenda today. We will go through why this requires your attention right now, where to look for AI opportunities, how to think when looking for AI opportunities, how to generate those high impact use cases, what makes the use case worth pursuing, how to go from vision to value, and also what's in store for the near term future, according to the work trend index.
We will also do a short demo for you to show some capabilities. And of course, summarize the key takeaways so that you can bring them directly back to your desk. So let's jump into it. So let's jump into it. First and foremost, why does this matter now? AI is no longer a future bet. It's a present imperative. CEOs and CXOs are under pressure to innovate, to optimize, to differentiate.
And AI is the lever and your best friend. According to the 2025 work trend index annual report, 82% of leaders say this is a pivotal year to rethink key aspects of strategy and operations. And 24%, and 24% say that their companies have already deployed AI organization-wide. And technology is shifting rapidly. We have already seen a major shift in the opportunities AI presents when moving from traditional AI focused on task automation and predictive analytics to generative AI.
Suddenly we have systems that can mimic human behavior, generate all sorts of text and images and personalize our interactions. And in this very moment, we're moving even further ahead towards agentic AI, where systems proactively identify opportunities, initiate actions, collaborate with humans. And we will see more of how that will affect us in the near term future later in the presentation.
But major announcements in tech and AI definitely comes at a much, much higher pace than ever before. And we do not know what the next big thing will be. But what we do know is that your data will never be ready for the next big thing, nor will your people. If you haven't already started experimenting and implementing use cases with generative AI and even agentic AI.
You can discover and ideate as much as you want and maybe even come up with the most disruptive use case there is. But even the best use case is useless if your people cannot build it, use it or govern it. People and culture are core enablers for AI transformation. AI literacy to AI fluency takes time and commitment.
And it's far easier to change your technology than to change your culture. So if you haven't already jumped on the AI train, it's time to go into discovery and ideation mode. So one way to start the discovery process is to look for business problems. Signals of processes that could benefit from automation or optimization with AI is, for example, repetitive rule based tasks.
Usually high volume processes that follow predictable patterns and consumes valuable human time. Data rich, insight poor areas where data is abundant but underutilized also signals opportunities for pattern recognition or summarization. Processes that rely heavily on expert judgment or manual communication, often slow or inconsistent and difficult to scale, are ideal for AI assisted decision support, drafting and knowledge sharing.
Paying painful or inefficient moments in the customer journey where responsiveness, personalization or clarity is lacking. It's also signals that indicates potential with AI and valuable information trapped in documents, emails or expert heads might be ripe for gen AI powered retrieval summarization or synthesis. Lastly, there's likely an opportunity to elevate customer experience and engagement through dynamic content, conversational interfaces or hyper personalization.
Another way of doing the discovery process is by flipping the script completely and reimagining how you approach problem solving. Instead of starting with a business problem and asking, can AI help us with this, you instead ask, what can AI do now that wasn't possible before and how might it transform our business, our product, our customer experience. With AI first thinking you start with capabilities, not problems.
Giving cutting edge AI capabilities, you ask, if this was a team member, what would we ask it to do? Secondly, you assume abundance, not scarcity. Traditional thinking is constrained by time and attention, but AI first thinking assumes infinite scale. What would we ask? What would you do if you had a thousand tireless analysts, designers, researchers? And lastly, think in systems, not in tasks.
Encourage designing AI augmented systems, not just automating individual tasks. So instead of automating invoice processing, imagine an AI that takes care of the entire vendor relationship lifecycle. And when we shift from traditional problem solving to an AI first mindset, we do not only do a technical pivot, but a strategic one. When we start with what AI can do rather than what's broken, we unlock ideas that leapfrog the competition.
It's not about incremental improvement, but about re-imagining what's possible. By aligning with where AI is going, not just where it is already, we build solutions that are resilient, scalable and relevant tomorrow, not just today. AI first thinking also nudges organizations out of cautious experimentation and into bold transformation. It encourages teams to test, learn and iterate, fostering a culture where innovation isn't a side project.
It's actually the default. And one of the most important aspects when ideating around AI use cases is understanding the relationship between human and AI. It's not just about what that AI can do, but how it fits into our workflows, our decisions, and even our values. We can think of this relationship as on one end, we have human led AI, where the human is fully in control and uses AI as a tool.
For example, a doctor using AI to support a diagnosis. And then we move into human AI collaboration, where both work together as partners, for example, a driver using a navigation system that is adjusting to traffic in real time. Further along. Further along, we see AI led humans where AI takes the lead and guides human actions, maybe setting priorities or assigning tasks.
And lastly, we reach independent AI, where the system operates autonomously without human intervention at all, like smart infrastructure, managing energy or traffic flows in a city. Now, here's the key message. The most successful AI use cases aren't just about technology. They're about augmenting human capability. So let's talk about a powerful metaphor from Microsoft research that helps us think about how AI impacts human capability over time.
It's called the steroids, sneakers and coach model. And it's a great way to evaluate the long term value of an AI solution. Steroids. These give you a quick performance boost, but at a cost. Think of an AI tool that automates responses in customer service. It might speed things up in the short term, but if the team becomes overly reliant on it, they may lose the ability to handle complex or nuanced cases.
Over time, skills erode, it's efficient, but it's not sustainable. Then we move over to sneakers. A good set of running shoes help you go faster, but they don't fundamentally change you. They support your existing abilities without making you better or worse. For example, using AI to translate survey responses or summarizing meeting notes. It saves time, but it doesn't necessarily build any new skills.
Lastly, we have the coach. And this is kind of the gold standard. A coach doesn't just help you perform better today. They help you grow. Imagine an AI that not only reviews your code, but explains why it suggested the changes. Over time, you become a better developer. And this is where AI becomes a long term enabler of human potential.
So as we design AI use cases, the question isn't just what can this do for us now? It's what kind of relationship are we building with AI and are we creating dependency or are we building capability? And now that we've explored how to generate AI use cases, the next step is figuring out which ones are actually worth pursuing.
That's where the AI use case radar comes in. This tool helps you prioritize by mapping each idea across two dimensions, business value and ease of implementation. On the business value side, ask yourself, does this use case meet a real customer need? Could it differentiate us in the market? Will it generate new revenue or reduce costs? And most importantly, does it align with our strategic goals?
Then on the implementation side, consider, do we actually have the data that we need? And is it high quality? Are there any privacy, security or legal hurdles that we need to take into consideration? How complex is the integration? And do we have the right people and skills to make it happen? The goal here is not to overanalyze, but to get a clear picture of where your quick wins are and which ideas might need more time, resources or risk management.
And it helps you move from a long list of possibilities to a focused, actionable roadmap. If you've identified a promising AI use case, the next step is, of course, to articulate it clearly. And this is where a business use case templates like this becomes incredibly useful. This template helps you move from idea to action by structuring your thinking around a few key questions such as how AI can help solve a problem, required stakeholders and business value.
And it truly lets you visualize if you have a strong AI use case or if you need to go back to the drawing board. This is also a helpful tool to have when aligning teams, securing buy in and ensuring that your AI initiatives are both impactful and feasible. Thank you, Ingrid. And as we move from vision to value, it's important to understand what are those phases, stages from strategic ideation to rapid prototyping, business app creation, risk and ethics review, and launch and scalable product.
So, as you can see, these five stages are resembling the software development lifecycle, because it's really tight. So, once that we have been discovered and ideate over one specific use case that we want to implement, we really need to adapt to what it needs to be done to implement the AI in this specific product or business application.
In this case, I will say the most critical that we have seen is, as Ingrid mentioned, the AI literacy. So, people really need to have the skills in order to realize the value that you have visualized during the ideation and the discovery stages to make real value on this. Some of the tools that you can actually show the value of this new business application or enhancing some of the capabilities of your already applications that the applications that already exist.
So, you can use some AI tools such as lovable. That is a tool where you can design a user interface in a matter of minutes. And this is built on AI, but this is just an example of how you can create value or rapid prototyping, creating a really fast forward user or final user feedback, loop feedback together with your product team to realize the value and assess that this is what it actually, the final user wants.
So, and this is just one example of how you can use AI into this cycle. But in reality, it should be embedded in all of these stages because that's what it will create a structure and efficient approach to this innovation within your company. But now that we have seen like, okay, now I have my use cases, I have already prioritized them.
I have really an strategic ideation on that. And now I am realizing the value or I'm in this process, like what is the path forward to this? So, for every leader and every employee, now it's the time to take decisive action and how to get started. So, we can divide this in three phases, but the first phase will be like human with a system.
So, every employee has an AI system that helps them to work faster and better. So, what does that mean? So, you can actually hire your digital or first digital employee to be ready to take actions and tasks that are traditionally done by humans that could be from answering support tickets to drafting reports. So, starting by this clear role or clear task, you can actually automate this task and it can actually add value to the job that you are doing every day.
So, and what it will be really like this new phase for you is like not just about efficiency, but about building like the workforce that blends your human creativity with AI unique strengths. So, in the phase number two, so it will be like define an AI team or human AI radio where you can actually have more than one assistant.
And then you can actually have like digital colleagues that are taking on specific tasks at human direction. So, this can be like small business processes that needs to be automated. And one of these, and each of these specific agents can help you and can help the team and the organization to be more efficient and reduce time, right?
So, like it can actually help you with some everyday work practices and also provide their skills into the ongoing learning as also as helping the human teams to reshape their tasks and be even more efficient. And in the third phase, humans set the direction and human agent and agents execute business processes and workflows, checking as needed. So, this will be the third phase when really humans are leading these agents.
And remember, this is the path forward. This is what the 2025 World Trend Index annual report is predicting as that that is really happening. So, that's why I opened this slide like this is now is the time to take intensive action. If you want to go forward advance in that as an individual and also as a company.
So, this is the things that we are seeing as a near future. And then, okay, I have this information, I know I need to create or to add AI, how I will scaling AI as a business capability. So, again, as an Ingrid mentioned earlier, this is more about people. This is more about the culture that we want to build into the as a business capability, and it will enhance or it will create a strategic differentiator for each of the company, even as a personal level, how you are adopting.
So, first, when it's the awareness phase, we are seeing like positioning AI as a strategic enabler is not just a technical tool. So, it's usually AI is seen as a tech initiative, but it's more than that. So, it needs to be in both cross-functional teams, for example, HR, finance, and the product development teams. So, it needs to be encouraging the experimentation, like the pilots, and tolerate the failure that is expected.
And then iterate fast from that. So, that will be like the adoption part of this. And then equip all the employees with tools and training to use AI. We have seen by experience that there is nothing more powerful to have really this training and access to AI. So, that is when individuals, like contributors use their own creativity to solve the challenges that they are facing daily basis, right?
So, this is when you are actually seeing the fostering of this creativity or innovation together with AI as enabled. So, that will be like the next one, the integration phase when you actually AI is embedded in the workflows. And you can actually build governance, ethics, and transparency into each or every single project that is embedding AI. And also making sure that this is aligned with your values as an organization.
And then, like the final step will be like taking into the leadership, right? In your, like, let's say your same industry. Because let's not forget that every company is all the time competing with some other competitors, right? So, this is what we believe AI will enable you as having a strategic differentiation when you already have it in place.
So, and what is the next step? So, what it will be to make this happen? So, we are not only talking here about what is like the future. Because there is some examples of companies that they are already at that level, right? And this is some examples of forward thinking companies and how they are using AI already.
So, for example, they still have the companies that built an agent to identify and consolidate consumer insights. So, instead of shifting through scattered reports and endless backups, teams can now pull up actionable intelligence instantly or almost instantly. So, buyer also is using like their R&D team is saving up to six hours a week when they are accelerating the development of products to drive innovation in agriculture.
Also, Wells Fargo, who actually deployed an assistant to financial services that it has like 3,000 bankers across 4,000 branch. So, now 75 of the searches have been through an agent. And this has been cutting times of from 10 minutes to 30 seconds. So, I will show you an example of how you can actually today like hire your first digital assistant and start in this journey.
So, I will just stop the presentation and I will show you, can you let me know if you see the right browser in red, please? No, not yet. Okay, let me just then switch the presentation. Okay. And then share again and window here. Can you see? Yes. So, this is just my Microsoft 365 co-pilot and I have created a briefing assistant.
So, I create this assistant with having in mind that I want to hide this assistant because I really need help to prepare briefings for my meetings. So, let me just find the prompt that I created for this example. So, in this example, I just use a very simple prompt. I will show you afterwards how I actually build this and how it's working in behind scenes.
But now let's use it. So, I create this prompt where I'm asking like create a brief for my upcoming meeting. The meeting will include the following companies and CEO representatives. So, I use like just a list from a Apple company with Tim Cook, Alphabet with Sundar Pichai, Microsoft with Satya Nadella, and it goes on in that list.
So, I'm assuming these people will attend my meeting. So, I will just send it and then it's instantly start to create an overview of each of these representatives. What is the recent achievements? What is the market position of each of these companies and people who is attending the meeting? And then it also creates some key discussion points.
So, what are the things that I possibly focus on during that meeting? Right? And generative AI, obviously. And then collaboration opportunities. So, in this fairytale meeting, right, where I will attend. So, it will provide me an insight of what it will be interesting for me to discuss in that meeting. But as you can see, my prompt is not really like really strong.
It's not asking too much. It's not providing too much context. But this is why I actually built this agent. Because I don't want to prepare the context every time. I don't want to create a really big or really detailed prompt every single time. And for that, I use, like, in order to create this assistant or this digital assistant, I use Copilot Studio low-code.
Which means I create everything, as you see, using the user interface without any code. Assuming that I have no authentication profile. But still, I can create or hide my first AI assistant. So, here, I actually create the description. Like, this is an agent that helps to prepare briefings for CEOs and CXOs. Focusing on recent progress and technology trends in generative AI analyst guidance.
And summaries of companies attending the briefing. So, then I provide specific instructions. For example, like asking, like, provide briefing anchoring recent progress. An analyst guidance. For example, from Garnet. Include a short summary. Like, ensure the briefing is concise, informative, et cetera, et cetera. So, I possibly could use knowledge basis from my own company. But I didn't use it in this specific example.
Even though there is some already -- I have access, but I didn't use it. And then I -- as you can see here in this started prompts. So, I can configure those prompts here. Based on each of the things that I want this assistant to help me. Like, I can come here and say, like, what are the latest trends in generative AI that I should be aware of.
And as you might see until this point, like, this is changing every single week. So, this for me, it's critical. So, but this is just an example of how you can actually start today to hire this, like, your first digital assistant. So, how you can really start to experiment and be conscious about what you can actually automate in your daily basis work.
Right? Like, just as simple as it is. So, this was just a very simple example. And let me just come back again to my presentation that, for some reason, I need to stop and go back to share. So, okay. I'm assuming I -- yes. So, with this, what are the takeaways of this presentation? Start small and think big.
So, pilot with a high impact and low risk use case. Don't just stay updated on the AI trend. Be bold on implementing them. Like, it's easy. Just -- you just need to start. Invest in AI fluency. Not just the tools, but the mindsets. So, if you are attending this session, it's obvious that you are interested in this topic.
Start today. Read a research paper. Do an example. Research. I mentioned one tool that is lovable. What is lovable? If you don't know that, just go on. Like, there is some smart pieces of information that you can -- it doesn't require too much time. Just start slowly, slowly. And, finally, thanks a lot for listening to our talk.
And, please let us know your feedback in the chat. Ingrid and I, we have been working with enterprise customers on their digital transformation journey for years. And, if this resonates with you, we would love to continue the conversation after the session. So, connect with us on LinkedIn and stay in touch. Thank you.
Transcribed automatically; names checked against the speaker record. Select a time to jump there.
Article based on this talk
How do you find AI use cases that are actually worth building?
Microsoft's Jessica Perez and Ingrid Aaseng share practical tools for discovering, judging and prioritizing AI ideas, and for taking the first step toward agents.
In short
Jessica Perez and her Microsoft colleague Ingrid Aaseng argue that finding valuable AI use cases is less about chasing technology and more about asking better questions and preparing people. Their talk offers two ways to discover opportunities, a model for judging whether AI builds or erodes human skill, and a radar for prioritizing ideas. It closes with a phased path toward human-led agent teams and a live low-code agent demo.
Key takeaways
Discovery and ideation
AI-first thinking
Use case assessment and prioritization tools
Agentic AI and the future of AI
Why people and culture come first
Ingrid Aaseng opens with urgency. Citing the 2025 Work Trend Index annual report, she notes that 82% of leaders call this a pivotal year to rethink strategy and operations, and 24% say their companies have already deployed AI organization-wide. AI has moved from task automation and predictive analytics to generative AI, and now toward agentic AI, where systems identify opportunities, initiate actions and collaborate with humans.
Waiting is not an option, she says, because neither data nor people will be ready for the next big thing without hands-on experimentation. Clever ideas alone do not carry a transformation: "But even the best use case is useless if your people cannot build it, use it or govern it." (3:55) Moving from AI literacy to AI fluency takes time and commitment, and she is candid about where the hard work lies: "And it's far easier to change your technology than to change your culture." (4:23)
Two ways to discover AI opportunities
The first route starts from business problems. Aaseng lists the signals that point to AI potential: repetitive, rule-based, high-volume tasks; data-rich but insight-poor areas; processes that depend on expert judgment or manual communication; painful moments in the customer journey; valuable information trapped in documents, emails or experts' heads; and chances to lift engagement through dynamic content, conversational interfaces or hyper-personalization.
The second route flips the script. Instead of asking whether AI can help with a known problem, teams ask what AI can do now that was not possible before. "With AI first thinking you start with capabilities, not problems." (6:06) She adds two more principles: assume abundance rather than scarcity (what would you do with a thousand tireless analysts, designers or researchers?), and think in systems rather than tasks. Rather than automating invoice processing, for example, imagine an AI that handles the entire vendor relationship lifecycle. As she puts it, "It's not about incremental improvement, but about re-imagining what's possible." (7:59)
Steroids, sneakers or coach?
Ideation also means deciding how humans and AI should relate. Aaseng describes a spectrum: human-led AI (a doctor using AI to support a diagnosis), human-AI collaboration (a navigation system adjusting to traffic in real time), AI-led humans (AI setting priorities or assigning tasks) and independent AI (smart infrastructure managing energy or traffic flows).
To judge long-term value, she shares the steroids, sneakers and coach model from Microsoft Research. Steroids deliver a quick boost at a cost, like customer service automation that slowly erodes a team's ability to handle nuanced cases. Sneakers help you go faster without changing you, like translating survey responses or summarizing meeting notes. The coach is the gold standard, such as a code review tool that explains its suggestions so you become a better developer. "A coach doesn't just help you perform better today. They help you grow." (11:32) The design question, she says, is whether a use case creates dependency or builds capability.
Choosing what to build: the radar and the template
Once ideas are on the table, the AI use case radar maps each one on two dimensions. Business value asks whether the idea meets a real customer need, differentiates the company, generates revenue or reduces cost, and aligns with strategic goals. Ease of implementation asks whether the data exists and is of high quality, whether privacy, security or legal hurdles apply, how complex the integration is and whether the right people and skills are in place. "The goal here is not to overanalyze, but to get a clear picture of where your quick wins are and which ideas might need more time, resources or risk management." (12:59)
Promising ideas then go into a business use case template that captures how AI solves the problem, which stakeholders are required and what business value is expected. Aaseng presents it as a way to align teams and secure buy-in.
From vision to value, and a first digital assistant
Jessica Perez takes over to describe five stages that resemble the software development lifecycle: strategic ideation, rapid prototyping, business app creation, risk and ethics review, and launch as a scalable product. She names AI literacy as the most critical factor and points to tools such as Lovable, which can design a user interface in minutes and speed up feedback loops with end users.
Drawing on the 2025 Work Trend Index, she outlines three phases: every employee working with an AI assistant, teams where digital colleagues take on specific tasks at human direction, and finally humans setting direction while agents execute business processes. Scaling AI as a business capability runs through awareness, adoption, integration and leadership, with cross-functional ownership, tolerance for failed pilots, training and access for every employee, and governance, ethics and transparency in every project. She cites examples including Bayer, whose R&D team saves up to six hours a week, and Wells Fargo, where an assistant cut bankers' search time from 10 minutes to 30 seconds.
Her demo shows a briefing assistant built in Copilot Studio without code, using a description, instructions and starter prompts so a short request produces a full meeting brief. "But this is why I actually built this agent. Because I don't want to prepare the context every time." (29:25) Her closing advice is to start small, think big, pilot a high-impact and low-risk use case, and "Invest in AI fluency. Not just the tools, but the mindsets." (32:40)
Watch the talk
"Discovery and Ideation of AI Use Cases" by Jessica Perez (with Ingrid Aaseng), 34 minutes: https://www.womentech.net/video/discovery-and-ideation-ai-use-cases
About the speaker
Jessica Perez is a Technical Specialist at Microsoft and a technical advisor specialized in product development, artificial intelligence and cloud architectures, with more than 18 years of experience across Latin America, Western Europe and the Nordics. She co-presented with Ingrid Aaseng, Cloud Advisor and Azure Specialist at Microsoft, who advises large organizations on digital transformation and innovation in the cloud.
Questions this talk answers
Where should organizations look for AI use cases?
What is AI-first thinking?
How can you tell whether an AI solution builds or erodes human capability?
How do you decide which AI use cases are worth pursuing?
How can someone start experimenting with AI agents today?
Drafted with AI from the talk transcript. Editor review pending.
Frameworks and ideas mentioned
AI-first thinking
Steroids, sneakers and coach model (Microsoft Research)
AI use case radar
Business use case template
Five stages from vision to value
Resources mentioned
2025 Work Trend Index annual report
Lovable
Copilot Studio
Cite this talk
Jessica Perez. "Discovery and Ideation of AI Use Cases." WomenTech Network, 2025. https://www.womentech.net/en-au/video/discovery-and-ideation-ai-use-cases