PressW brought in team members with the most experience in the healthcare specific AI space to help with this due diligence. We met with the team at Ecliptic to understand their investment criteria and generated a question set from our 5 pillars of AI competency. We then interviewed the healthcare company’s team directly to complete the picture on the AI aspects of their company.

Below are the areas of interest for our team:

Data Quality and Selection

We not only evaluate broadly for the selection of data sources that are representative of the broader population of data that the AI solution could see, but also ensure proper measures are taken to ensure that data is high quality and reliable.

Data Governance and Responsibility

This category evaluates the responsible storage and usage of data. Is the company securing their data responsibly? What measures have they taken to reduce bias and set up their models to make fair and equitable predictions across various demographics?

AI/ML Modeling Sophistication

We evaluate the AI/ML technology used, along with it’s specific applicability to the task at hand. Oftentimes, nuanced and crafty application of these technologies is more important than using the most bleeding edge model you can get your hands on.

Testing and Evaluation

Here, we evaluate all the mechanisms that the company has in place to evaluate model performance against their core objectives. This segment also includes the actual results their models are achieving.

Solution Scalability

Many AI solutions seem promising in the POC phase, but Deloitte finds nearly 70% of AI solutions don’t make it to production. Here, we evaluate the scalability potential of their solutions from POC to production and onward.

From the calls and application of this framework, we understood at a deep level how this company was approaching their AI future and distilled our findings into a report that we shared with the Ecliptic team. Given our team’s background in the VC investment space we were able to point out a few flags both good and bad from a tech perspective that contributed to Ecliptic’s final investment decision.