At the beginning of the course, I thought about working on a disruption hypothesis in the logistics and supply chain industry. I have experience as a warehouse employee, so I know some struggles employees face, such as repetitive and unstimulating tasks, heavy manual labor, especially during peak seasons, and distress (of new employees) from the high demand for speed and accuracy. I talked to my groupmates about the possibility of implementing something like an Augmented Reality (AR) smart glasses in warehouses, which I took inspiration from the professor’s previous project. Not only will it improve their accuracy and efficiency, it may also potentially reduce employee churn rates which is a timely issue in this industry. Aside from this idea, another fellow member thought about tackling the healthcare industry, primarily regarding the administrative workflows involved. The whole group agreed that both ideas had potential so we decided to pitch both ideas to our professor and ask for her opinion. Based on what the professor said and her reaction, we decided to push through with the logistics industry. I later did some quick research just to see if the idea is really worth writing about. Much to my surprise, some companies have already set the idea in motion. This led us to push with the healthcare industry’s administrative workflows. When my fellow member further explained his proposed topic, I honestly thought that it was far from my interests and also way out of my field of expertise/knowledge. But it was, without a doubt, a timely and high-value problem that must be addressed somehow.
For the short summary or proposal of our project idea, we divided the task evenly so that everyone will have an equal share of work. I was assigned to write the initial value logic, which was specifically to determine for whom the solution creates value and how the organization would capture value. From my research, I learned that this solution will create functional and economic value for the entire healthcare ecosystem as it will streamline the administrative procedures and reduce costs. For this part of the project, we didn’t critique each other’s work because we saw that each member contributed equally. We also judged the short proposal as well synthesized, as if a single person wrote the summary when a group of 4 worked on different parts independently. So, when we received the feedback, we were all caught off guard. We thought we’d made our point and that we were off to a good start, but the feedback said otherwise. The problem was on the broadness, where the proposal was very abstract and not describing the specific workflow we want to focus on. We appreciated the critique because it allowed us to ponder and eventually narrow it to one administrative workflow of healthcare professionals in the rehabilitation center. We were also advised to do this so that the prototyping and evaluation phases will be easier for us.
Before we started planning how to approach the project, the group thought it might be best to ask relevant professionals in the field such as doctors, nurses, or administrative staff about bottlenecks they encounter at work. As a group, we were lucky to have a member that works closely with these professions so the interviewing or gathering of observations was smooth sailing. When enough observations were gathered, we were then able to determine what administrative workflow we wanted to focus on. None of the project members also had experience in making working prototypes so we consulted a professional who works closely with Agentic AI. We reached a final decision on Careflow, an EHR overlay that reduces the administrative burden on rehab centers’ healthcare workforce. As it is an overlay on existing EHR systems in the organization, it will not require the organization to discontinue using their legacy systems. We believed this is financially more advantageous for the rehab centers since they may have already invested a significant amount on the legacy systems and are in essense difficult to replace. The core value of this system is in its ability to reduce the amount of time for healthcare providers to check, analyse, and review individual patient documents across platforms.
My main tasks in this project was to determine the business need and challenge, business model and recommendations. First, for the business need and challenge, I wrote about the problem, the targets, solution, what happens if no action is taken, and the need to solve the problem now. Second, I created a business model canvas encompassing the proposal’s value creation, delivery, and capture. Third, I developed recommendations on what to implement, change, scale, and test next. Lastly, I helped on other parts mainly for formatting and synthesizing sections like the AI Canvas and the possible risks. As a group, we created the slide deck in a way that best presents the information we’ve gathered. While researching and putting the slide deck together, each member reviewed others’ sections to ensure cohesion and consistency.
The collaboration not only was a way for me to review key concepts from the lectures but it also enabled me to learn a lot about how the healthcare industry operates especially regarding invoicing and how these organizations make profit in the end. The topic is rather niche so not a lot of people know of this system, but since it has become the topic we worked on, I did my best to learn everything I could to best support the proposed solution. However, if given another opportunity, I would try to be more proactive like completing all coursework and readings ahead of time.
Working on this topic changed my understanding of AI in business strategy because I was able to see how much impact it actually brings, how powerful it is and how it makes our lives easier in many ways. In our case, AI made it easier for the group to develop a prototype, as what used to take several hours or days now takes a few minutes and a few iterations. With just a complete and well-structured prompt, we can make a prototype for our demo day. To quote what one of our audience said during the demo day “It is a very extensive prototype.” Without AI, it would have been a struggle to develop such a prototype so I really do think highly of this technology. Yes, it does remove the barrier to entry on tasks like creating prototypes but it is still important to keep in mind the possible risks, especially when applying it to actual business implementation. Risks concerning data governance as well as legacy integration can be expensive and irreversible. Ultimately, it is not just about the ease of prompting or making prototypes with AI, but it is about how we manage and take into account the possible risks of using AI, especially in industries like healthcare, which is highly regulated.
