How Generative AI Puts a Twist on the Build vs Buy Decision
ERIK SWENSSON, VP SALES
Feb 07, 2024
The “should I build or buy” software decision has been one organizations have faced for decades. Companies must ask themselves, should they utilize their development efforts on an application or buy something already built for this purpose and focus resources on a different and more impactful problem. Typically companies put development ‘build’ resources on projects that off-the-shelf software cannot supply, where complex business or regional regulatory requirements exist and for items that will truly differentiate them from their competitors. Back-office, productivity, and security products typically fall into this ‘buy’ category.
Generative AI applications promise to bring new capabilities and efficiencies to businesses. While legacy off-the-shelf applications may offer new AI features, companies may still have the need to build tailored AI applications in-house to satisfy niche business needs. However, developing custom generative AI requires significant data science resources — which are in high demand and limited supply — alongside development resources. This scarcity puts a twist on the traditional build vs buy software decision. Data scientists are considered far more scarce than development resources. In fact, insufficient data science talent is a top barrier to the technology’s adoption, according to Anaconda’s “State of Data Science 2022 report”. Unlike traditional software, quality generative AI relies heavily on curated datasets, robust models, and continual tuning from experienced data scientists.
With the global shortage of data science talent, developing custom generative applications risks diverting valuable data scientists away from high-impact business projects. Attempting to build generics generative apps in-house could delay other data initiatives that more directly impact core operations and strategic goals.
Utilizing off-the-shelf natively designed Generative AI applications such as Digital Tutor and Knowledge AI for training, onboarding, and knowledge management can lower the cost of development and increase time-to-value. Companies can put to use built and tested applications like Procurement AI to build contracts and reduce procurement times. This will allow you to focus Data Science Efforts on high-impact or regulated areas within your business. Their time is best focused on developing generative applications that generate a long-term competitive advantage or enable wholly new business capabilities.
