top of page
Writer's pictureJenny Pollock

Build With AI Right Now

Article by Lindsay Witmer Collins, originally posted on the WLCM blog, The Journal. 




At WLCM, we’re full steam ahead on building AI apps for the OpenAI marketplace. This piece is Part 2 in a series helping entrepreneurs understand this new frontier. You can see Part 1 here and Part 2 here.


The other day, I logged on to LinkedIn and saw a video showing an AI voice commentating a soccer match. Predictably, the comment section was split between “This is stupid!” and “This will change everything!”


I often find myself asking why we tend to apply AI to the most human endeavors. Why are we creating AI that can write poetry instead of AI that can, I don’t know, do my taxes or explain my health insurance policy to me? 


The answer, I think, lies in a business problem that long predates generative AI. 

Business decisions and technological developments frequently unfold along a path divorced from the daily realities of the people actually doing the work. Innovation happens to them, instead of with them or for them. If you’ve ever been subjected to using software built by someone who doesn’t understand how you do your job, you know what I’m talking about.


For this reason, the novel-but-useless AI use cases populating our newsfeeds are often the result of imagination rather than necessity or even opportunity. 


This isn’t a bug. It’s a feature. Innovation typically starts in the tech world and takes some time to trickle out into its most useful contexts. I believe generative AI lays the groundwork for that journey to happen faster and allows a broader base of people to participate in its evolution.


I’m talking about you! 


Custom GPTs help bridge the knowledge gap between problems and their solutions. So what might AI look like for you and your product? 


AI you can implement right now

Recommendation algorithms

Few functions can be as delightful or helpful as a finely-tuned recommendation algorithm. They can help customers get the most out of the product and upsell new features and offerings — think Netflix’s “What to Watch Next” suggestions or your grocery store’s loyalty program.


Of course, most of the ads, emails, and recommendations we see are a total miss. If you can nail your recommendation algorithm, it will instantly set your product apart and earn the trust of your users. 


If your product has a library of resources, content, or products, and you’ve been wondering how to get users to explore or dive deeper into that library, a recommendation algorithm might be the answer.


Virtual assistants and experts

How often have you wished you could just take your brain, or someone else’s, and put it into your product? With AI, it’s more than possible. 


AI experts help users better navigate the “unknown unknowns” of the task at hand. It can help them understand the “why” of what they’re doing and keep them from making missteps. Consider a financial app that provides specific guidance for users to stay on track toward their goals.


I also think of the efficiencies an AI expert could offer when injected into an internal business tool. For example, AI experts could accelerate new employee training by helping them adopt on-policy practices and workflows in addition to building their general knowledge. As more tenured employees retire, AI experts can help retain their valuable institutional knowledge for the next generation.


Scheduling and planning systems

For many types of businesses, scheduling is a moving target. How do you schedule employees when your staff is thin? When a customer cancels their appointment at the last minute, how do you source a new customer and save that revenue? How can you prioritize the most profitable jobs and customers? 


AI-powered scheduling and planning systems can make these adjustments on the fly and at scale. This capability is often a game-changer for businesses, both by reducing manual work and capturing more revenue.


Chatbots

Let’s face it, the state of website chatbots and customer service call trees is dismal. Up until now, these tools have been restricted to a rigid decision tree model that rarely proves helpful in the face of the variety and nuance of user needs.


AI chatbots work on large language models that allow customers and the chatbots to communicate in organic, natural language. This creates a more person-like helper in tone, experience, and helpfulness. 


Putting the pieces together  

One of WLCM’s clients, Brella, operates three childcare centers in and around Los Angeles, and they’ve created both internal and customer-facing apps to help the business run smoothly and profitably. 


Brella’s goal is to ensure each center is operating as close to full capacity at all times, and we accomplish that through machine learning. Machine learning allows the app to determine availability and implement a dynamic pricing model based on that availability. When centers are at capacity, users can get a place on a waitlist, so when cancellations happen — and they do, often at the last minute —Brella’s app decides who gets the newly vacant spot and notifies that customer of the new availability. 


Generative AI allows us to add a new layer of intelligence by learning from our data and making suggestions. We can ask it what we could have done in the past to optimize profit and experience, and the generative AI can make observations that inform what we do in the future. Not only that, we can ask the AI to play out different models and scenarios to see the implications of the path we choose. 


AI can make observations about customers through their behavioral data that allow Brella to stay ahead of the curve. For example, the AI could tell us the markers of a customer who is likely to cancel their appointment. For Brella, this heads-up can inform a more seamless availability notification process.


For others, it could help with retaining customers or spotting particularly desirable or undesirable behavior. The applications are endless, and OpenAI’s GPT store will make it easier to layer on these functionalities.


Reaching real time

Up until now, it’s been difficult to create technology that can observe and change in real time. Compiling data, analyzing it, and changing the software based on those insights has taken so long that by the time the changes go live, the situation has shifted in one way or another.


AI promises to put businesses ahead of the technological curve and make available the knowledge that will inform impactful change. 


The human challenge is to understand and align around what matters most deeply and which problems are worth solving.

1 view0 comments

Comments


bottom of page