How does Artificial Intelligence and semantic search work for a portal

Sure this is easy. You'll require a massive amount of cash. If you have to ask "how much cash", then you won't explore AI. Once you've determined your AI tech, then you'll go through a multi-year process of Knowledge Acquisition, where you extract knowledge of your entire manual process from every person engaged in the current manual process. Begin training your neural nets. After a few years... to a decade... you'll have a Poe'esque (Ala "Altered Carbon") level AI to do your graphic design for you... auto-magically... I remember going through this process at MCI Telecommunications years ago. Fairly straight forward. Just took $1,000,000s of hard cash over a 10+ year period to convert their switching tech to AI'esque style logic.

One of the most innovative uses of AI that I've come across recently is using it to analyze sales calls. They're able to break down conversations into digestible/searchable topics. From here, they analyze what phrases and topics lead to/away from conversions.

I am happy to volunteer! I have a more "classical" background in mathematical statistics, and I have followed the distribution of artificial intelligence with great interest and curiosity; also because, I feel, that the actual application and real life-value of AI-implementations often drowns in technical aspects. So what is actually delivered is more the Data Scientists "wet dream" rather than something, that works in real life. If you are interested, feel free to contact me. Best regards Kenneth Wolstrup

Clinical Decision Support coupled with cost optimization will grow over the next ten years in light of AI and Amazon development. So what do I mean in a practical sense by this? Say you are a patient with abdominal pain. Your doctor thinks you may have a gallstone, but it could also be a kidney stone or perhaps a bulging disc in your back among other conditions. A test would help your doctor diagnose you. But what test? It could be an ultrasound, CT, MRI, or a nuclear medicine study along with blood work. But what would be the best test for you given your medical history, past tests you've undergone and cost of the tests in the area? This is way too much information for a busy doctor to analyze in a 15-minute office visit. Ahh, but there is an AI platform with all your past medical history and costs of local services Amazon has figured out. Your doctor, working in conjunction with the platform can enter relevant clinical information based on their physical exam and interstation with you to determine the optimal test to order to diagnose your condition. This would result in high-quality, cost-effective care to be delivered with the help of AI in collaboration with your doctor's input. How does this change the job of a doctor? Well, it would make them more efficient and able to provide higher quality. They would need to learn how to interact more closely with an AI platform and recognize a stronger partnership with technology. As a result, decrease mental burden and job satisfaction would result. Ultimately, there could be lower rates of provider burnout and increased job satisfaction with more individuals wanting to enter medicine. This could help address our issues around limited access in light of an aging population requiring more services.

I have worked with a number of AI tools - both webbased and locally installed tools - and have a background in mathematics/statistics. So I am familiar with this space. My suggestion is first and foremost focus on the use case - forget that it is an AI tool. If your tool provides value, and solves problems for your customers, they will come no matter the underlying technology. When you say "web driven", I assume that this is some kind of subscription model, and you don't expect to sell anything face to face? Again, I would focus on use case (we solve problems), value creation (how much do we save or make?), ease of use (just upload this file and watch the magic happen), cloud installation (no need for own maintenance or local data scientists to maintain models). In the AI-space, your competitors are often internal employees - data scientists, statisticians, etc., and they are familiar with their own tools. So find out, who is your target segment in the businesses. Most likely a data scientist would tell you, that "I could build that myself". But what about following maintenance cost, depending on individuals, etc. So do you target to data scientists or business users? Often successful projects happen in collaboration; one can not live without the other. With regards to your focus on the US I am curious, why you focus there. Is that where you are located physically? Or there you see the most likely target customers? Since you are web driven SaaS, couldn't you go anywhere? I hope this helps. Feel free to schedule a call, if I can help you further with this. Best regards Kenneth Wolstrup