InsightfulDiscussion

1208: Quality, Trust, and Scale | Cal Brouilette, CFO, Flatiron Health

CFO THOUGHT LEADER46m 12s

Cal Brouillette, CFO of Flatiron Health, discusses his 20-year career spanning energy, technology, and healthcare, and how Flatiron balances the tension between scaling AI-powered data extraction, maintaining quality and trust in oncology research, and driving sustainable business growth.

Summary

Cal Brouillette traces his career from a three-year graduate scheme at Centrica in the UK to his current role as CFO at Flatiron Health. His early move to the US at age 26, transition from energy trading to Silicon Valley technology, and subsequent roles at Salesforce and Zendesk prepared him for leadership in a healthcare technology environment. He joined Flatiron as head of strategic finance six years ago before assuming the CFO role.

Flatiron Health operates a two-sided business model: on one side, it provides oncology-focused electronic medical records (EMR) software to community oncology clinics that administer more than 50% of cancer care in the US; on the other side, it sells real-world evidence (RWE) datasets and analytical services to pharmaceutical and biotechnology companies. The company was acquired by Roche but operates independently, a trust-building effort that took years and remains central to the business.

AI represents Flatiron's most significant technology transformation. Previously limited to analyzing 500,000 patients due to labor-intensive manual curation, the company now leverages large language models to extract and structure unstructured medical data, expanding to 5 million patient records. This enables the company's new "Panoramic" data product line and an emerging agentic layer called Flatiron Telescope, which answers scientific questions in real time rather than requiring days of analyst work.

Brouillette emphasizes that AI is a powerful tool but not a panacea. While finance teams can now receive AI-generated answers to complex questions in minutes, the central challenge is trust. He describes building a robust data warehouse over two years before deploying AI on top of it, allowing him to understand the underlying data well enough to validate AI outputs. The company is still developing audit frameworks and confidence protocols to operationalize AI-driven decision making at scale.

The strategic tension animating Flatiron's planning is balancing three forces: quality (maintaining gold-standard RWE), trust (with customers, partners, and regulatory bodies), and scale (expanding data scope and datasets globally). Growth is the top priority for the coming 12 months, requiring decisions about investment levels in AI, network development, and technology infrastructure. Brouillette's focus as CFO is ensuring decisions are tracked and validated rather than simply executed and forgotten.

About this episode

<p>Cal Brouilette recalls a sleepless night in Houston, speaking at 3 a.m. with someone from Blackstone about a purchase-and-sale agreement and transition-service agreements. It concerned a power-plant portfolio sale at Direct Energy, a Centrica subsidiary, that Brouilette places in 2013.</p><p>According to Brouilette, the assignment arrived in late September, when the CEO wanted the portfolio sold and announced before year-end. That left at most 120 days. Brouilette says he was managing finance and accounting, functioning as a business-unit CFO and loading up on MBA classes at Rice because his job had seemed stable.</p><p>The new assignment changed that calculation. Brouilette says he ran its finance work “soup to nuts,” covering the valuation taken into negotiations, management presentations, purchase-and-sale terms, and transition-service agreements.</p><p>The transaction presented him with what Brouilette describes as a $700 million “go or no go” decision. “Yes, this was hard,” he recalls thinking, but the experience prompted this realization: “I can do this job. I can make these decisions.”</p><p>Brouilette says the moment gave him confidence to make high-stakes decisions quickly while recognizing that he could “still keep learning.” He describes it as his shift from operator to strategist.</p><p>At Flatiron Health, Brouilette says his focus as CFO is not to “make a decision and let it ride,” but to consistently track results before the organization travels too far down “the wrong path.”</p><p>His remarks present strategic confidence not as certainty, but as the discipline to decide, keep learning, measure results, and correct course.</p>

Key Insights

  • Cal argues that moving from energy to technology required bridging a credibility gap, as technology leaders in Silicon Valley underestimated the sophistication of skills developed in mature industries and traditional finance structures.
  • Flatiron Health's expansion of AI-extracted patient data from 500,000 to 5 million records was not driven by having more patients, but by AI making previously cost-prohibitive manual curation economically viable while maintaining quality standards.
  • Cal contends that the core challenge with AI in finance is not speed or efficiency, but trust—he piloted AI tools for three months and found they produce answers 90% reliable, but establishing audit frameworks and confidence protocols for operational decisions remains a work in progress.
  • Cal claims that Roche, as a 10-20 year cycle thinker, has enabled Flatiron to invest more thoughtfully in long-term technology and network development than venture-backed technology owners typically would, despite initial concerns about independence when the acquisition was announced.
  • Cal argues that cancer treatment is increasingly fragmented into narrower patient populations, making larger and more diverse datasets essential for identifying patient cohorts relevant to specific research questions.
  • Cal describes a pivotal moment in 2013 negotiating a $700 million power plant sale in 120 days while enrolled in an MBA, which convinced him he had the judgment and confidence to make high-stakes strategic decisions as a CFO.
  • Cal asserts that Flatiron maintains its credibility as a gold-standard real-world evidence provider by refusing to trade quality for speed, even as AI enables faster data processing—this constraint shapes all technology investment decisions.
  • Cal contends that the finance team's ability to trust AI outputs depends entirely on pre-AI investments in centralized data platforms and data architecture, not on AI tools themselves, making infrastructure investment a prerequisite for AI value.

Topics

Career progression across energy, technology, and healthcare sectorsFlatiron Health's two-sided business model and market positionAI and machine learning applications in healthcare data extractionBalancing quality, trust, and scale in oncology data productsRoche acquisition and maintaining independence and credibilityReal-world evidence (RWE) and its market valueFinance team transformation through AI tools and data platformsDecision validation and measurement discipline in strategic planning

Transcript

Support for CFO Thought Leader comes from Salesforce. Unify selling and billing for a seamless customer experience. And OneStream. Trusted data. Faster decisions. Hello, this is Martino Cadorni, CFO at Depot, and you are listening to the CFO Thought Leader podcast. This is episode 1208. Yeah, so the graduate scheme in the UK that I was on as part of Centrica is a three-year program. You do three or four rotations. You qualify and study to become a chartered accountant in the UK at the same time. And the third rotation is actually funny. It just came up, I think it was 20 years today. I was in Portugal when I found out I got this role in the US…

Full transcript available for MurmurCast members

Sign Up to Access

More from CFO THOUGHT LEADER

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.