Three Questions to Vet AI Vendors
By Keith P. O'Brien
4 minutes estimated time to read
One of the striking things about Artificial Intelligence (AI) is its seemingly unlimited potential--it can be applied to an array of challenges. There is no shortage of higher ed challenges that AI could help tackle given the competitive pressures, funding issues and vast data repositories on campuses. My earlier blog suggested 6 foundational questions to help campus leaders establish their approach to AI. However, for administrators, identifying the right problem for AI to target and picking an AI platform inevitably involves working closely with vendors; AI doesn’t lend itself to a DIY strategy!
As with any disruptive technology in the marketplace, a thriving AI vendor community or ecosystem (to use business speak) exists. Crunchbase (tracks technology startups and companies) tracks 9,986 companies in the AI space. Undoubtedly more and more consulting companies and technology firms are promoting their AI platforms and expertise to two- and four-year institutions. Here are three key questions to help administrators find an appropriate AI partner.
THREE QUESTIONS TO VET AN AI VENDOR
- What Does Your Technology Do? Given the specific challenge(s) our institution seeks to address, how does your AI platform/model pinpoint the student, processes, or systems that specifically influence the outcomes we want to achieve? If you have never tackled this issue with a college or university, do you have examples from analogous consumer-focused industries like financial services or health care?
- How Do We Use It and How Difficult Is It to Use? How does the AI translate its insights into recommended actions for campus units and students? How are these recommendations communicated to staff and students? How does the platform access and integrate the data within our CRM, SIS, LMS and communication channels? Institutions should also consider:
- Colleges and universities that adopt an AI-enabled initiative are, de facto, empowering an algorithm to shape decision making. Therefore, understanding how the algorithm works and decisions will be operationalized on campus are important
- If implementing vendor recommendations requires introducing new channels (for example, conducting campus wide texting) then investigating the cost, time and privacy implications are crucial
- How Do You Manage Data Privacy? As every AI algorithm requires massive data sets which specific data sources do you need from our systems? For all student data do you need any personally identifiable information (PII)? And how do you maintain FERPA and GDPR compliance? How do you share the new data sets generated by your algorithms with us?
These questions are start starters--raising some of the core issues that apply irrespective of the type of institution or challenge. To gain a bit more insight on vendor offerings and selection, take a look at these articles:
- Classifying the AI Vendor Landscape: Making Sense of the Market: a 2019 Forbes piece that helps demystify the world of AI vendors
- 6 Tips for working with AI vendors: a set of tips for CIOs that are applicable for higher ed and non-CIOs!
This blog is based on our free eBook about AI's game-changing potential, Artificial Intelligence and Higher Ed: Cutting Through the Noise
To learn more contact: Info@discourseanalytics.com