Ensuring AI Due Diligence for a Private Equity (PE) Firm
In this AI due diligence case study, learn how Clarkston Consulting partnered with a Private Equity (PE) firm to ensure integrity and quality in their AI due diligence processes. Read a synopsis of the project below or download the full case study.
Download the AI Due Diligence Case Study Case Study Here
Clarkston recently executed a project reviewing and auditing the codebase for an artificial intelligence (AI) product and the tangential software infrastructure for a private equity (PE) firm that invests in growing healthcare companies.
Prior to engaging Clarkston, the PE client was exploring a potential new investment in a medical device company that makes wearable electrocardiogram (ECG) heart monitors and utilizes AI to analyze the ECG heartbeat images 24/7 in a full-disclosure transmission and cloud-based platform. The client recognized that not only was the AI that the medical device company developed at the core of the company’s overall value, but also that most of the intellectual property surrounding the AI and other associated software was dependent on only a small handful of highly technical developers on staff. For that reason, the PE client saw it necessary to partner with Clarkston to have the AI product thoroughly evaluated for quality, vulnerability, and scalability and to receive recommendations on potential improvements.
Clarkston conducted a thorough, line-by-line evaluation of the AI-related scripts, with key focus across the company’s Convolutional Neural Network (used for image classification) as well as across data extraction and transformation scripts. Specifically, the scripts utilized outdated Python coding libraries which were no longer maintained, exposing the product to security risks. Additionally, other foundational challenges became apparent, such as production-level code that had insufficient documentation describing what each section of code was trying to accomplish, evident best-practice flaws with file paths, and a variety of unnecessary library dependency imports adding memory demand to an already memory-intensive process.
The Clarkston team also conducted a high-level review of the surrounding software infrastructure, including the patient portal, billing applications, the software that runs the actual devices, and more.
Clarkston worked with the PE firm and the target company, providing a thorough report outlining critical challenges and opportunities, as well as their overall risk level and a potential path towards resolution. Additionally, given the high concern around documentation, we provided the client with an extremely detailed framework for proper documentation of the AI development, testing, production, and maintenance process.
This project exemplifies the added value Clarkston brings to PE clients by enhancing the thesis and providing conviction into the investment. Clarkston’s work empowered the client to proceed with clarity, knowing their investment in this innovative medical device company is well-founded and poised for future success, while also ensuring there was a rigorous assessment of the AI codebase and surrounding infrastructure.
