Why Starting Small With AI Produces Lasting Results with Keisha Downes | Podcast Shorts from ArceeHQ

In this clip

Keisha Downes shares the hardest lesson from her organization's AI rollout in clinical documentation integrity. Leaning too heavily into probability-based technology before fully understanding its limits created real friction early in implementation. The fault, she reflects, lay neither with the vendor nor her team but with the decision to scale too fast before the system had demonstrated consistent, reliable performance.

Key Takeaway

AI tools built on probability can never be 100% accurate. Starting at smaller scope, measuring actual performance, and scaling only after demonstrated results is how organizations avoid early setbacks and reach a productive, sustainable collaboration with the technology.

“I should have started slow and really worked our way up.”
Keisha Downes, Vice President of Middle Revenue Cycle, Beth Israel Lahey Health

InnovationLeadership

From the full episode

Listen to the full conversation.

Keisha Downes

Season 2 · Episode 3 · 34 min

Data-Driven Prioritization, Responsible AI, and Operational Transparency in Revenue Cycle Leadership with Keisha Downes

Keisha Downes · Vice President of Middle Revenue Cycle, Beth Israel Lahey Health

Keisha Downes shares how data-driven prioritization shapes middle revenue cycle strategy across a multi-hospital system, from setting the roadmap to selecting and implementing AI tools. Drawing from her experience at Beth Israel Lahey Health, she walks through a clinical documentation integrity initiative involving a third-party AI tool that scrubs 100% of inpatient encounters, the lesson of starting small rather than over-relying on probability-based technology, and how case mix index serves as the key metric for tracking impact. The conversation covers her view of the path from reactive to proactive denial management through agentic AI, the AI governance framework her organization uses to evaluate vendors and prevent tool sprawl, and her daily practice of monitoring the candidate-for-billing report to clear barriers before they affect AR targets. Keisha also shares her leadership philosophy of radical transparency, her advice to not let perfection be the enemy of good, and how cultivating curiosity in emerging leaders shapes the future of the profession.

Spotify Apple YouTube

Listen to the full episode

Go deeper

From the clip to strategy.

The themes our guests cover are the same ones our executive guides and ROI calculators are built around. If this clip resonated, here’s where to take it next.

Claims Denial Management, powered by agentic AI

How ArceeHQ catches denials before they happen, resolves them automatically, and gives your reviewers only the cases that need a human.

See the solution

Executive Guide

Reducing claim denials with agentic AI

A practical framework for revenue cycle leaders. Covers the operating model shift, what to measure, and where automation actually pays.

Claims denial ROI calculator

Model the revenue impact of a first-pass rate improvement at your health system. Plug in your denial volume and payer mix.

Run the calculator

Request a demo

Let’s redefine financial engagement together.

See how ArceeHQ can supercharge your revenue cycle. We’ll walk through the platform, discuss your specific systems and goals, and show measurable lift you can expect in the first 90 days.

Reply within 24 hoursNo spam, ever