The good and the bad of off-the-shelf AI
AI solutions aren’t all that different from investments—there are plenty of options, discernable levels of risk, and ample room for growth in AI adoption, but every organization has a custom portfolio built for its specific needs. Building your own AI models isn’t for everyone. Every financial service copy has its own expertise, capabilities, and resources that help dictate how comprehensive its automated intelligence process should be. In many cases, that process includes hiring outside vendors and adopting off-the-shelf AI solutions. For the purposes of this conversation, we’ll designate AI vendors as a third-party company that either builds models or sells off-the-shelf models for consumption in your core business processes and requires your organizational data as a feed.
In my previous post, The most critical decision in building out enterprise AI: Build in-house or bring in a vendor, we explored the highs and lows of building out AI in-house. To continue that theme, today’s conversation explores seven reasons, some good, some bad, why companies hire AI vendors and how different financial service companies have used these partnerships for better or worse.
3 bad reasons to hire an AI vendor
- You can’t hire a good data scientist
Without a doubt, AI demands a certain level of expertise to implement properly. With a significant worldwide shortage in data science workers, it’s understandable why organizations would reach out to a vendor to fill that gap. Companies might not know where to start, don’t know what questions to ask potential candidates, or might not understand how to identify qualified talent. Instead, it becomes easier to outsource the problem and let someone else deliver the solution.
By failing to develop their own data science capabilities, companies set themselves up to be indebted to outside organizations. While most AI consultants do honest and amazing work, there are plenty of horror stories of partnerships turned sour and brands left with a stack of code and data that they don’t know how to operate on their own.
- Data science isn’t your specialty
Brands regularly bring in AI vendors because they think the vendor can deliver better than they could themselves. The practicality of this approach is fair, but the mindset behind it can be crippling. Data science has become table stakes for the future of financial services. Whether they represent a private equity firm, a national bank, or a community credit union, companies that adopt data science into their services give themselves a significant competitive advantage.
This mindset traps organizations in the past and hinders efforts for innovation. In order to stay ahead of the competition, businesses must take the initiative to form their own data science teams.
- You want to learn from the experts
Many organizations will bring in an AI vendor with the intention to watch and learn from their expertise. The company will observe as the vendor executes their own processes and methodologies, then try to recreate them on its own. Unfortunately, data science isn’t a recipe. Vendors build their necessity around their ability to obfuscate their efforts; they don’t want to train themselves out of a job. Organizations aiming to reverse-engineer a vendor’s data science capabilities often find themselves frustrated and overwhelmed. By all means, you should absolutely learn from a vendor’s results, but don’t expect to gain huge insights from their process.
3 good reasons to hire an AI vendor
- The vendor has data or tools you cannot recreate
Some vendors have access to unique datasets or methodologies that companies simply cannot reproduce themselves. These vendors specialize in providing unique industry datasets that are far more comprehensive than anything else on the market.
- The vendor has a different perspective
Occasionally, the right AI vendor brings a new perspective to your team. The ability to fold in fresh eyes to a challenge can lead to fantastic learnings for your organization. Partnering with an insurance specialist to build policy risk models, for example, can help understand your business in a new and exciting way. Remember that most of these vendors have seen problems similar to yours across a wide breadth of situations. Transplanted learnings from that could prove to be a boon to your processes.
- You need a synergistic solution
Vendors work best as an extension of your team. Whether intending as a stop-gap solution or a complementary team, AI vendors can help an in-house analytics team manage new challenges, scale effectively, or fill a temporary need.
Buying an off-the-shelf AI solution should be a viable resource for any growing financial services company—or any other organization for that matter. By bringing AI into the financial services conversation, companies have the opportunity to expand their capabilities, understand their customers, and drive transformation across the industry.
If buying an AI solution doesn’t interest you, visit my previous post and explore the peaks and pitfalls of building your own automated intelligence models, or email me directly at [email protected]. I’d love to hear your story!
