Cornerstone AI: Making real-world data clean, actionable, and reliable
By Vishal Lugani, Asad Khaliq, and the Acrew Team
Acrew led Cornerstone AI’s latest round of funding. If you’re a biopharma company, a life sciences data commercialization business, or a services vendor to biopharma, Cornerstone AI is a must-have partner for you. They are fast becoming the neutral, trusted way that biopharma makes data actionable, by leveraging AI to clean real world healthcare data significantly faster and more accurately than other solutions.
Underappreciated in healthcare is that there now exists a path to strong health data availability. A decade after the EHR became widely adopted, and several years after regulation passed to drives interoperability and data transfer standards in the US healthcare system, rich health data is accessible to those that have a legitimate need.
Despite these advances, the data is rarely actionable (i.e., data quality and integrity has a long way to go). There is a burning need we’ve observed among life science companies that use and buy that deidentified data to assess the quality of the data they buy or even generate quality improvements themselves, because they need data:
Quality Checked: rapidly characterize the quality and completeness of every dataset in an automated consistent manner for comparison across data sources
Cleaned: find errors in the data, correct them if possible
Standardized: convert all free text entries to industry standard medical terms (why does my lab data have 150 ways to represent neutrophils?)
Cohorted: collate all patients who meet study criteria without needing to understand the details of the data schema from a larger dataset
Harmonized: combine multiple data sets and de-duplicate so the power of a larger dataset can be leveraged (why does every data provider use a different schema?)
Auditable: explain how raw data gets to final data in a way that would pass regulatory scrutiny (FDA is encouraging RWD submissions but is holding them to similar standards as clinical trial data)
This is really hard and is not just a technical problem. Cornerstone AI has pulled this off a depth of data science, AI, life sciences, and regulatory expertise. Having gotten to know Michael Elashoff over the course of a year, we became convinced that he had assembled the team for the job. Almost everyone has worked together previously in highly relevant roles for several years. Mike, Andrew Howland, and Clara Oromendia, as co-founders, brought decades of biostatistics, genomics, and data science expertise to bear. The team has strong engineering capacity and a grip on when AI is needed, ML is best, or data science is the best path to solving a problem. They understand the regulatory constraints pharma & life sciences face. With the addition of now CEO Viraj Narayanan, who experienced the problems associated with unclean data throughout his career and has a track record of scaling life science businesses, we believe Cornerstone has remarkable leadership and team maturity for their company stage. And customers agree!