AI in Action at Berkshire Partners

By Jon Nuger , Managing Director, Berkshire Partners and Limor Gultchin , Vice President of AI and Machine Learning, Berkshire Partners

Across industries, artificial intelligence (AI) has emerged as a key lever of competitive advantage. According to RSM’s 2025 Middle Market AI survey, Generative AI (GenAI) adoption has surged to 91% among middle market companies, up from 77% last year — a clear sign that AI is becoming standard in business operations.

At Berkshire Partners, we see AI as an urgent and transformative opportunity. As a multi-sector specialist investor in private and public equity, Berkshire has long embraced advanced analytics and data science as part of its value creation toolkit. That foundation has enabled us to rapidly expand into applied AI across our portfolio, our investment process, and our internal operations.

AI enablement at Berkshire underpins our long-term approach to value creation. We view AI as complementary to our investment principles, providing new ways to scale operations, unlock insights, and transform business models. Across our portfolio, companies are at varying levels of maturity—from early experimentation to operating at scale—but all recognize its potential.

Our framework rests on three pillars of enablement:

As our portfolio companies rapidly adopt and deploy AI, we are already seeing it drive both operational efficiencies and bold new strategic initiatives. It is delivering value across the portfolio—from reimagining customer service and automating deliverables to empowering sales, marketing, and software development with smarter tools. No matter the sector, from healthcare to technology, the common thread is a forward-thinking mindset—teams are moving fast to realize the real-world impact of AI’s potential.

Since mid-2023, Berkshire’s AI working group has convened CTOs, CIOs, and heads of AI from across our portfolio to share strategies and execution models. These sessions, featuring over 15 vendors and industry leaders since their inception, have not only introduced new tools into team workflows but have also fostered a culture of cross-functional insight-sharing and responsible governance.

To help teams identify high-value AI opportunities, we encourage them to map their value chain and ask where AI can reduce friction or create leverage. Typical starting points include:

• Manual, repetitive, and time-consuming tasks, especially those at scale
• Underutilized structured or unstructured data that could inform decisions or automate workflows
• Unstructured documents—such as contracts, reports, or emails—that can be mined, summarized, or co-authored
• Historical decision-making records or SOPs that can guide and assess AI model behavior
• Labor bottlenecks or specialized knowledge that is hard to scale
• Workflows with tolerance to some level of error, human review loops or variation in output

Prioritization is critical. We advise companies to assess opportunities on both business impact and technical feasibility, often using a 2x2 evaluation. Many teams begin by tackling two to three “low-hanging fruit” use cases while also placing one or two big bets that could deliver outsized strategic value.

One standout example of Berkshire’s AI enablement is Ensemble, a portfolio company that manages revenue cycle operations for hundreds of hospitals across the U.S. Ensemble is accountable for more than $45 billion in net revenue across more than 30 healthcare systems and has established itself as an industry leader in applying AI to reduce friction, improve financial outcomes, and elevate patient and clinician experiences.

Ensemble built a broad AI roadmap spanning multiple initiatives, from automation in claims processing to predictive analytics that strengthen clinical documentation. The example that follows represents just one of Ensemble’s many robust AI initiatives underway, illustrating how the company is embedding AI across revenue cycle operations. Among these initiatives, one of the most impactful has been Ensemble’s GenAI platform for clinical alignment, designed to address a pervasive industry challenge: payer denials. These denials, which reject hospital claims for reimbursement, are a significant source of lost revenue and an administrative burden. Clinical payer denials are rejections of hospital claims by insurance companies, typically due to questions about the medical necessity of care or accuracy of clinical documentation. These denials create significant financial and administrative strain for providers, often requiring a complex appeal process to recover owed revenue.

Here’s how the denial-management solution works:

• Automation at Scale: The platform rapidly processes payer denials and translates thousands of pages of clinical criteria, industry guidelines, and patient details into persuasive clinical arguments tailored to effectively appeal each case.
• Smart Drafting: It automatically generates customized appeal letters in seconds, then routes them to clinical experts for review, validation, and submission.
• Human in the Loop: Clinicians oversee outputs, focusing their expertise on complex cases rather than routine paperwork.
• Predictive Insight: By analyzing over 80,000 denial audit letters, Ensemble is piloting predictive modeling to prevent denials upstream, strengthening clinical documentation before claims are ever submitted.

The impact has been substantial. According to Becker’s Hospital Review, Ensemble has accelerated denial appeals by 40% and improved overturns by 15%. Clients also report significantly lower final denial rates: just 2.2% of net revenue, compared to an industry average of 2.8%.

This success underscores Ensemble’s philosophy: AI is not a side project but a core enabler of transformation. The denial-management platform is only one of many initiatives in its robust AI roadmap, aimed at creating a frictionless revenue cycle where providers recover revenue they are owed and can focus on patient care.

Berkshire sees Ensemble’s success as emblematic of its broader AI philosophy: combine cutting-edge technology with deep operational expertise to empower teams with tools that scale their impact, while keeping people in the loop where judgment and nuance matter most.

Several factors contributed to Ensemble’s success:

• Deep industry expertise ensured AI models were grounded in real-world clinical and payer contexts
• Cross-functional collaboration between technical teams, operational leaders, and clinical experts kept solutions practical and scalable
• Human oversight balanced speed and efficiency with the accuracy and credibility necessary in healthcare settings
• Continuous learning from tens of thousands of denial letters built predictive models that not only address denials reactively but also prevent them proactively

AI is no longer a “future technology.” It is here, now, shaping the way businesses operate, compete, and grow. Berkshire Partners is committed to ambitious but accountable AI enablement. Our goal is to empower our portfolio companies to adopt AI responsibly, creatively, and with measurable outcomes.

Ensemble’s experience is just one example of the many AI initiatives currently underway across our portfolio—and just one of many within Ensemble itself.

Stay tuned for our next article where we’ll highlight another example of applied AI impact in the Berkshire portfolio.

This article reflects the subjective views of Berkshire Partners LLC as of the date of this article only, which views are subject to change. Individual examples of companies are illustrative and are not intended to reflect all companies in which Berkshire has invested. Nothing herein constitutes an offer to sell, or solicitation of an offer to buy, any security.