Blog: The Clinicians Who Started Going Home on Time
Editor-in-Chief
What three behavioral health organizations learned when AI moved into the note, the UR queue, and the compliance check.
When Banyan Treatment Centers began rolling out AI-assisted documentation, Cara Bishop, the organization’s VP of Clinical Services, saw almost immediate benefits. The early pilot cut note-writing time by 30 to 42 percent, and as the rollout grew from two sites to seven, that figure climbed past 60 percent. Cara was enthusiastic about those numbers, but they weren’t the real win for Banyan.
That came when clinicians reached out to say that, as a result of the AI, they were going home on time, satisfied they’d done a positive day’s work instead of spending their evenings revisiting sessions.
That thread ran through the entire Elevate 2026 panel on AI for clinical productivity, moderated by Kipu’s Sally Abu-moustafa. Panelists from a nine-state treatment provider, a justice-focused residential campus in Phoenix, and an 80-clinic opioid treatment network kept arriving at the same place: the time savings are real, but the retention story may matter more.
A compliance bar that keeps moving up
Cara had set what she assumed was an ambitious first-pass compliance target of 85 percent, and her teams cleared it within three days. She raised the bar to 90, and that is where it lives today. Clinicians hold their own notes until they hit the threshold, and supervisors check the score before anything else. Cara frames that step as verification rather than surveillance, because a team that can see the tool getting it right learns to trust it.
At New Freedom, Chief Clinical Officer, Katherine Nisbet, leads a peer-driven workforce serving roughly 400 residents coming out of incarceration. Many of her staff are new to behavioral health, and what kept her up at night were notes with incomplete sentences and no documented service. AI gave those staff a way to find a clinical voice without losing their own.
The revenue cycle win nobody expected this fast
Within three days of giving her utilization review team an AI chart assistant, Cara’s organization secured authorization on three cases they had fully expected to be denied, which added up to 29 days of residential treatment. Her larger point landed hard: payers are already using AI to review your documentation, so “good enough” no longer holds. Documentation has to be unimpeachable.
The CEO’s caution
Nick Stavros, CEO of Community Medical Services, framed AI as process improvement in the lineage of Lean and Six Sigma, but warned that you cannot outsource your thinking. His biggest gains have come from data analytics, and he treats every output the way a good supervisor treats a first draft, asking for the research behind a conclusion, then the research against it, then why the model landed where it did.
He also named the gap between how executives and clinicians see AI. Leaders think in ROI, while frontline staff think in what he calls return on effort, and closing that gap has been his hardest work.
Monday morning
Katherine’s advice was blunt. Your staff are already using AI, with or without a policy, so give them a sanctioned place inside the EMR and explain the why. Her message to her teams is simple: we’re getting AI so you can go home on time.
Watch the full Elevate 2026 session, AI Strategies for Clinical Productivity, Operational Efficiency & Treatment Outcomes, to hear how each organization approached governance, what an AI policy looks like across 80 clinics, and how the loudest skeptics became ambassadors.
Want to see how Kipu can help your organization?
Request a personalized consult below.
About the Authors
Rely on Kipu to keep you ahead of change.
Subscribe to Kipu for behavioral health news, updates, community celebration, and product announcements.