Product descriptions:
Ensemble manages one of the largest healthcare data footprints in the U.S., supporting hospitals and health systems nationwide as they navigate rising payer pressures and razor-thin margins. Before implementing Databricks, the company struggled with disconnected SQL Server databases, gaps in tracking patient journeys, and labor‑intensive data pipelines — issues that hampered large‑scale insight generation and slowed innovation.
Today, with the Databricks Data Intelligence Platform — powered by Databricks AI and Lakebase — Ensemble has unified more than 2 petabytes of provider, payer, and clinical data into an AI-ready asset. This foundation unlocks predictive modeling, real-time insights, and agentic AI workflows that help providers recover more of the revenue they’ve earned, faster and more reliably. What wasn’t possible before — proactively flagging denied claims, automating AR workflows, and democratizing access to enriched, low-latency data — is now driving measurable impact across the healthcare ecosystem while laying the groundwork for a next-generation, AI-native revenue cycle.
Fragmented data and rising payer pressures put provider margins at risk
Ensemble is the leading end-to-end revenue cycle management (RCM) partner for hospitals and health systems across the U.S., helping providers maximize yield and ensure accurate, efficient reimbursement for the care they deliver. Their mission is to keep healthcare organizations financially strong so they can reinvest in patient care and their communities.
Unlike point solutions, Ensemble uses 2PB of harmonized data and AI-driven insights to reduce revenue leakage — catching underpayments, preventing denials, and optimizing reimbursement strategies before claims reach a payer. Their single goal: predict where revenue will be lost and fix issues proactively, so hospitals recover more, faster. A key innovation is agentic AI that predicts payer behaviors. Ensemble can flag likely denials — even when payers indicate otherwise. “We can prevent the denial before it ever happens,” said Grant Veazey, CTO, Ensemble. The result: cleaner claims, fewer denials, faster collections, and freedom from costly appeals.
According to Grant, “We can now predict, no matter what the website says or what the person on the phone tells you, when you need a pre-approval. We can prevent the denial before it ever happens.” This predictive capability drives cleaner claims from the start, minimizes denials, and accelerates collections. Ultimately, this frees providers from the costly cycle of appeals and resubmissions. Before Databricks, Ensemble’s growth was constrained by significant data silos and technical limitations. Each new customer meant setting up a separate database with its own schema and enrichment rules, making it impossible to generate insights across clients. “Like other RCM technology platforms in the industry, our insights were limited to one client at a time. We had 15 individual SQL Server databases with 15 individual schemas. Our insights were limited to one client at a time. We had no visibility across clients until we could harmonize that data.”
These same gaps made AI difficult to operationalize. Without a unified, harmonized dataset, models trained on fragmented client-specific data couldn’t generalize across customers. Furthermore, legacy systems lacked the scalability to handle training at a petabyte scale — forcing teams to downsample data and limiting model accuracy. Deployment was equally challenging, as manual pipelines and siloed environments meant models couldn’t be consistently promoted into production or monitored at scale. As a result, Ensemble struggled to move beyond proof-of-concept AI projects, with models too fragile and costly to deliver a reliable business impact. What they needed was a modern data platform capable of scaling with their rapid growth, unifying disparate data sources, and unlocking advanced analytics and AI-driven insights.
Ensemble recognized that maximizing value on complex patient journeys requires a unified approach to integrating clinical, pharmacy, lab, and radiology data alongside administrative data from EMRs. Streamlining data pipelines accelerates valuable insights and supports rapid claim recovery. As Chief Architect Dragon Sky explained, “Every time I have an engineer working on a server or debugging infrastructure, that’s time they’re not spending understanding the data and how we can get value from it.”
Powering end-to-end revenue intelligence with Databricks AI
Ensemble adopted the Databricks Data Intelligence Platform — a unified foundation for managing, enriching, and operationalizing their data at scale to break free from fragmented systems and manual workflows. The shift enabled Ensemble to move from isolated SQL Server databases to a modern lakehouse architecture, harmonizing data into a single, AI-ready asset. “The hard work we’ve put in over the last three or four years to build out this platform in the right way, with Databricks as a partner, has put us in a really unique position,” said Grant.
By standardizing raw data in the bronze layer of the medallion architecture, preparing and harmonizing it in the silver layer, and enriching it in the gold layer, Ensemble created a single source of truth across clients in Delta Lake. This makes it possible to reconcile financial performance “to the penny” on a transactional basis, while providing consistent, high-quality data to downstream analytics and models.
A cornerstone of Ensemble’s innovation strategy is Databricks AI, which powers its agentic AI systems. These systems orchestrate multiple models — including custom-trained models, Cohere, OpenAI and Claude, accessed through AI Gateway and Amazon Bedrock — to handle different tasks across the revenue cycle. With Databricks AI’s integrated capabilities, such as Model Serving for deployment, AI Gateway for secure external model access, and monitoring for usage and cost optimization, Ensemble can unify external and internal models within a single framework. This allows them to scale AI consistently, maintain governance, and continually refine performance. “Everyone is excited about AI and agentic in particular. As we've done evaluations, Databricks has really shown through here,” said Dragon. “With the platform being serverless and fully managed end-to-end, we don’t have to manage — or even think about — all the bits and bobs, but we can still do that evaluation loop really well.”
With MLflow, Ensemble can log, compare, and deploy multiple models simultaneously. They can experiment with Cohere, OpenAI, Claude, and open source alternatives while maintaining cost efficiency. “I like quickly switching out a high-cost model and evaluating a cheaper one that still gives us 90% of the performance for a fraction of the cost. I really think that will be the magic that everyone will need.” This flexibility allows Ensemble to balance accuracy, cost, and speed as they scale agentic AI across their revenue cycle workflows.
Another key enabler is Lakebase, a fully managed Postgres database integrated with the lakehouse. Lakebase serves as the transactional data layer, connecting harmonized data directly to applications and AI agents, thereby eliminating the need for complex data pipelines. With Lakebase, teams can materialize views on demand and deliver insights faster. According to Dragon, “We used to have to build these manual data pipelines. Having Delta tables as a baseline makes it easier and simplifies things, so you don’t need all the manual work.”
This has been transformative for Ensemble’s accounts receivable (AR) follow-up workflows, where predicting, prioritizing, and resolving payment issues drives massive value for healthcare providers. By connecting real-time payer and provider data into materialized views, AI agents can quickly analyze patterns, flag anomalies, and recommend the next best action items. “Lakebase lets an agentic team quickly self-serve the data they need for their models — whether it’s historical claims or real-time transactions — and that’s really powerful,” said Dragon.
With Lakebase, Ensemble has successfully democratized data access. Self-contained pods of data scientists, engineers, and AI specialists can iterate faster. “Instead of throwing requests over the wall to data engineering, each pod can self-serve and create the needed views. It makes teams fully self-sufficient,” explained Grant. Lakebase helps turn siloed, complex healthcare data into AI-ready intelligence, enabling Ensemble to develop smarter, faster, and more proactive solutions that directly improve provider financial outcomes.
Improving revenue yield by up to 5%
Ensemble has transformed how health systems capture, protect, and grow their revenue by unifying more than 2 petabytes of provider, payer, and clinical data on the Databricks Data Intelligence Platform — curated and enriched over the past decade. With predictive models, low-latency insights, and AI-ready infrastructure, Ensemble helps hospitals reclaim revenue that would otherwise be lost, at a scale equivalent to the second-largest hospital system in the U.S.
“We’ve been able to drive upwards of a 20% efficiency across our own organization over the last couple of years,” said Grant. “And for our customers, we’re consistently seeing a 3–5% net revenue lift — well above industry averages.” Ensemble’s performance is validated by industry analyst KLAS, which awards them a score of 95 —the highest ever achieved in revenue cycle management and nearly double that of others in the industry.
This uplift isn’t just incremental in an industry where most providers operate on razor-thin 1% margins. A 3% increase in a $1 billion health system translates to an additional $30 million in recovered revenue — money that can be reinvested directly into patient care and community health initiatives. With its harmonized data platform, Ensemble is uniquely positioned to respond faster than individual providers can on their own. It can detect denial patterns early, adapt AI models in near real-time, and scale AI-driven decisions across entire healthcare systems.
For Ensemble, the partnership with Databricks isn’t just about efficiency gains — it’s laying the groundwork for a next-generation, AI-native revenue cycle. Through Databricks AI and Lakebase, Ensemble is exploring new ways to automate decision-making, empower non-technical teams, and surface insights that were previously invisible. From predicting payer behaviors to optimizing entire patient journeys, Ensemble is building the intelligence layer providers need to thrive in an increasingly complex, AI-driven healthcare landscape.