BHB VALUE: Tackling Staffing and Reimbursement Challenges: The Role of AI-powered Automation in Successful Revenue Cycle Management

This article is sponsored by Thoughtful AI, based on a discussion with Kathrynne Johns, CFO of Allegiance Mobile Health, and Alex Zekoff, CEO of Thoughtful AI. This discussion originally took place on March 13, 2023, during the BHB VALUE Conference. The article below has been edited for length and clarity.

Behavioral Health Business: Could you talk a little bit about the growth challenges that behavioral health providers face today?

Kathrynne Johns: Billing processes in behavioral health tend to be very antiquated. Each payer sets unique expectations based on their individual agreements, introducing more steps into the billing routine and creating complications. Further, there is a high demand for medical billing professionals, and the talent pool isn’t large enough to meet it, so automation will be a key tool for organizations looking to expand swiftly and efficiently. The shift towards automation is inevitable.


Alex Zekoff: At Thoughtful AI, we build Fully Human Capable (FHC) AI Agents to help providers like Kathrynne solve this issue. Our Agents turbocharge these processes and workflows to address constant staffing and reimbursement challenges so that providers can focus on the patient experience. We are highly convinced that AI and automation are the answer to these systemic issues, not only in behavioral health but across the healthcare system at large.

Could you talk a little bit about staffing shortages and staffing constraints and their impact on RCM efficiency?

Johns: Especially in the medical billing sector, the workforce is predominantly older women who are often stuck in their ways of tried-and-true coding and billing practices. Attracting younger individuals or anyone willing to engage in the detailed, repetitive, and time-consuming nature of billing tasks is increasingly challenging across many sectors. Additionally, when hiring new staff without prior experience, the initial training investment can pose a significant cost for providers.


The process of integrating a competent and qualified employee can span from three to six months, varying with the number of insurance carriers involved and the intricacies of billing regulations. In contrast, automation through technology, such as using bots for billing processes, is a significantly more efficient solution. Once programmed, a bot can perform tasks consistently, whether five or 5,000 times, without absences, health issues, or turnover. This reliability, efficiency, and accuracy make automation a compelling option for streamlining billing operations.

As we look ahead at the next wave of the talent pool, how do you see administrative and billing staff evolving?

Zekoff: Behavioral health is an $80 billion industry growing 5% a year, and it is expected to reach $120 billion in the near future. The increasing hiring rate of RCM staff is growing linearly with revenue right now, and deploying AI is critical to turning that growth into profit.

Suppose you have 25 people in your RCM department handling eligibility, claims, and payment posting. If you aim to double your growth, you may need to double your staff. However, imagine not having to hire 25 more people and avoiding the challenges of staff turnover. When you deploy an AI Agent, it turbocharges your existing staff so they become superhuman, eliminating concerns about turnover. The challenge of recruiting the next generation for these roles becomes less pressing as the younger generation, particularly Gen Z, shows little interest in these positions, opting for different career paths instead. Despite this shift, the demand for these jobs persists. The solution lies in leveraging AI to augment these roles, viewing it as an innovative approach rather than a problem.

How significant is the reimbursement and denial issue in behavioral health, and what are some of the main causes of that?

Johns: Denials often come from incorrect claim submissions. Navigating behavioral health claim submissions is notoriously complex — it’s like the Wild West, with each insurance carrier and state setting unique requirements for modifiers and codes. In turn, proper authorization and credentialing are fundamental to successful claim processing.

The introduction of AI Agents improved this process, and my first collaboration with Alex showcased the impact, as we saw over a 10% increase in the collection rate for claims within the first month of implementing bot technology. This improvement was primarily due to the bots’ ability to submit cleaner claims, streamlining the meticulous review of codes, credentials, and authorizations — tasks that are otherwise laborious and repetitive.

What’s the relationship between staffing levels and reimbursement results in healthcare generally?

Johns: Most of the time, staffing for claim processing is managed based on the ratio to the volume of claims. Typically, there are a specific number of billers to handle eligibility checks, others to manage the pre-bill process, and additional staff to address follow-ups on rejections, denials, and payment posting. These staffing needs directly relate to the monthly claim submission volume. The introduction of bots, however, fundamentally changes this dynamic. Unlike human staff, a bot can handle the process no matter how many times it needs to be repeated, making the volume of claims irrelevant to staffing requirements. This shift represents a significant advantage in managing staffing levels for Revenue Cycle Management (RCM).

What are some practical ways to tackle staffing and reimbursement issues in healthcare RCM simultaneously? Can you tackle both things simultaneously?

Johns: I believe so because once you put AI Agents in place, then your expectations for your team change. You have bot managers instead of people who are doing the tedious, time-consuming process, and you can dial in what functions your team is responsible for. Their job description changes.

Zekoff: You would think people might be fearful of AI coming in, but we’ve found that staff members who embrace this technology love their AI Agents because they reduce the strain and workload. If you have to go through 100 clicks every time you process a claim, it can literally cause mental health issues and burnout. People don’t want to do this work anymore, and this is not where our human self is best.

Our human selves are best when interacting with each other, solving problems, and using creativity, not when serving as a human-like augmented processor to move data. AI integration is good for humanity, and it’s good for the bottom line. All providers should think about this as they look to invest in AI in the long term.

How long would you say it took to cover the cost of the bot, and when do you think cost savings equaled the investment that you made in the technology?

Johns: COVID-19 forced me to reduce my team by half, as we were providing autism services for children, which wasn’t easily adaptable to telehealth. Once we started experiencing growth again, the process of hiring and re-training staff put a significant burden on our team, so I partnered with Alex to introduce bots into our workflow. We started with five, and the return on investment was immediate. The bots proved their worth, not just in terms of cost but also in speed. With a setup time of 90 days, this was significantly shorter than the training period for new hires, eliminating concerns about turnover.

Before COVID-19, we had two staff members handling pre-billing. After, we managed with just one, supplemented by bots, and we even increased our claim submissions. The bots also enabled us to quickly catch up on two weeks’ worth of billing that had been deferred at the year’s start due to deductibles, accomplishing it in just one day. We couldn’t have done it with people.

Zekoff: Our model operates on an AI-as-a-Service basis. We aim for a payback period of less than 4-6 months and handle all maintenance and optimizations. Implementing an AI Agent costs the same as hiring a person, yet it performs 10 times the amount of work. When considering long-term ROI, you can expect an 8-10 times return, while in the shorter term, it’s around 3 times.

What are the use cases for which behavioral health providers use AI and automation?

Zekoff: Eligibility verification, claims processing, payment posting, prior authorizations — those will get you to the core KPIs of reduction in DSO, denied claims and everything else. Additionally, coding, note review, revenue reporting, and reconciliation encompass the entirety of full-stack RCM. Expanding into finance, some of the use cases include accounts receivable, accounts payable, and payroll processing for HR. We have developed a bot capable of automating processes across the healthcare sector, and provided it meets specific criteria that apply to virtually any system, it’s straightforward to implement our bots. We also have a pilot program in which we deploy a bot in 45 days so providers can test and see the magic firsthand. We begin with RCM and can then expand throughout the organization.

How does Thoughtful ensure that the implementation of AI and automation in RCM aligns with maintaining a positive patient experience throughout the healthcare journey?

Johns: By accurately billing and following the RCM process, we can manage to dispatch clean claims, secure payment from insurance carriers, and process those claims correctly. Then, we’re significantly contributing to a positive patient experience. Enhancing our ability to send out more clean claims with the help of the bot means we’re offering our patients the best service possible.

Zekoff: From a patient experience perspective, we’re looking at no-surprise patient responsibility invoices for thousands and thousands of dollars. It’s all about accuracy- error-free claims going out of the door. AI Agents don’t make mistakes unless you train them to, and that simply means there is a problem with your process to begin with. Providers are swamped right now, and it’s difficult enough to prevent internal errors, let alone manage errors from patients and payers as well. AI solves that problem and creates a better patient experience, long-term relationships, and overall improved service.

Editor’s note: This article has been edited for length and clarity.

Thoughtful AI deploys AI across your healthcare organization to maximize profitability and unlock operational excellence. To request a demo and learn more, visit

Companies featured in this article: