1. What motivated your organization to participate in this year’s Global AI Awards?

Healthcare is one of the areas where AI has tremendous potential, but the value has to be practical. We wanted to highlight a project where GenAI was not introduced as a separate tool for people to learn or another step to add to an already complex workflow. It was embedded directly into an existing process to help people do the work correctly the first time. The result was measurable improvement across both operations and the revenue cycle. That combination—useful AI, thoughtful integration, and tangible business impact—is exactly the kind of work Sphere wants to contribute to the broader AI conversation.

2. Could you give us an overview of the AI solution or breakthrough you submitted for consideration?

Sphere developed a GenAI-based medical documentation validation service for a medical practice that was struggling with claims being returned for missing or incomplete documentation. Revenue Cycle Management staff were spending significant time manually reviewing cases, and corrections were delaying reimbursement. The solution integrates into the practice’s existing AdvancedMD EMR workflow and reviews medical documentation in real time against standardized claims requirements. If required information is missing, Medical Assistants receive feedback while they are entering the documentation—before the case moves further downstream. The impact was significant: cases returned to Medical Assistants for additional documentation decreased by 65%, average reimbursement time improved by 20%, and daily outstanding accounts receivable decreased by 35%.

3. How did your team collaborate to develop and refine this AI innovation?

The project began with the workflow rather than the model. Sphere worked closely with the medical practice’s staff to understand the different claim types, required documentation, and common reasons cases were delayed or returned. Together, we standardized those requirements so the AI had a clear and consistent framework against which to evaluate documentation. From there, the team identified where an automated review would provide the greatest value and designed the GenAI validation service around those points in the existing workflow. The solution was thoroughly tested for accuracy and reliability, and Sphere trained the client’s internal team to manage it after deployment.

That collaboration was critical. The technology had to fit the way the practice actually worked, not force the practice to reorganize itself around the technology.

4. What impact do you expect your AI work to have on the broader AI community or society as a whole?

We think this project illustrates an important opportunity for AI in healthcare: moving intelligence upstream.

Traditionally, incomplete documentation may not become visible until a case reaches another reviewer or a claim encounters a problem. By then, someone has to find the missing information, return the case, make the correction, and restart part of the process. GenAI made it possible to identify those issues while the documentation was still being created. That principle extends well beyond this particular use case. Many healthcare administrative workflows involve unstructured information, complex requirements, and expensive downstream rework. AI can help people catch problems earlier without removing them from the decision-making process. Done well, that means less administrative friction and more time available for higher-value work.

5. Were there any notable challenges during the development of this AI solution, and how did you overcome them?

One of the biggest challenges was turning complex insurance claim requirements into a consistent validation framework. Different types of claims require different information, and the reasons for rejection or delay were not necessarily captured in one simple rule set. Sphere first analyzed the practice’s workflow, claim types, required data points, and common sources of delay. We worked with the practice to standardize those requirements before introducing automation. The second challenge was integration. The solution had to provide useful feedback in real time without disrupting the Medical Assistants’ existing AdvancedMD workflow. That meant focusing as much on when and how feedback appeared as on the AI itself. Extensive testing and staff training helped ensure that the solution was reliable, understandable, and manageable by the practice’s own team after implementation.

6. How does your organization nurture a culture that drives continuous AI innovation?

Sphere’s mission is to expand human potential through AI, and that shapes how we approach innovation internally and with clients. We encourage experimentation, but we also expect AI initiatives to answer practical questions: What problem are we solving? How will we measure success? What data and controls are required? And how will this work in production? That mindset keeps innovation connected to outcomes rather than novelty. Sphere combines AI engineering with evaluation, governance, data readiness, security, and human-in-the-loop design so teams can test ideas quickly without treating production concerns as an afterthought.

We also believe the people closest to a workflow need to be part of designing the AI around it. Some of the best innovation happens when domain expertise and technical expertise meet around a clearly defined problem.

7. What advice would you offer to teams or companies aiming to make meaningful contributions in the AI space?

Start with a point of friction where improvement can actually be measured. In this project, we did not begin by asking how GenAI could be used in healthcare. We began with a specific problem: too many cases required additional documentation, manual review was consuming valuable time, and reimbursement was being delayed.

We would also recommend standardizing the underlying process before automating it. AI cannot compensate for requirements that no one has clearly defined. Understand the workflow, establish what “correct” looks like, decide where human judgment belongs, and then determine where AI can make the process faster or more consistent.

Finally, measure the business outcome. A successful AI system is not simply one that produces technically impressive output. It should change something that matters.

8. What are your organization’s long-term goals in AI, and how do you plan to advance the field moving forward?

Sphere’s long-term focus is helping enterprises move AI from experimentation into dependable production environments where it can become part of everyday work. That means continuing to advance areas such as agentic AI, knowledge and retrieval systems, document intelligence, AI-powered software engineering, and enterprise governance. It also means improving the infrastructure around AI evaluation, observability, security, human approval, and data readiness because those capabilities determine whether an AI system can safely scale beyond a pilot.

Healthcare is especially important in that work. There are substantial opportunities to reduce administrative burden and improve operational workflows, but the standard for reliability, privacy, governance, and accountability has to remain high. We believe the next stage of AI will be defined as much by how responsibly systems operate as by what models can technically do.

9. Are there any emerging AI technologies or trends your team is particularly excited about right now?

We are particularly excited about the convergence of document intelligence, multimodal AI, and agentic workflows.

A huge amount of enterprise work still begins with unstructured information, medical documentation, claims, invoices, contracts, forms, images, and conversations. AI is becoming much better at interpreting that information in context rather than simply extracting text from it. The next step is connecting that understanding safely to operational workflows. That is where agentic systems become especially interesting. Instead of only identifying an issue, AI can help route it, gather supporting information, prepare the next step, or escalate it to the right person. Combined with strong evaluation, auditability, permissions, and human approval, that creates the potential for AI to become a dependable participant in complex enterprise workflows rather than simply an assistant sitting beside them.

To dive deeper into Sphere's award-winning work, visit their website at https://www.sphereinc.com