Balancing the decision between Horizontal or Vertical AI SaaS applications
Businesses globally are finding more and more ways for Gen AI to support their processes, products and services. Financial Institutions are no different.
As Financial Institutions consider how to deploy AI, they face two initial questions. Where amongst the 100’s of processes do they start? And then which tool do they use? This is before they get into the practicalities of making it work within their existing workflows and alongside/within their existing tooling.
The first question is a matter of business prioritisation and assessing where will the lowest investment deliver the greatest scale and impact. The second question is less obvious.
Existing AI tools fall into two broad categories, horizontal and vertical. Both types offer value when deployed against the appropriate type of task, either generalist or specialist. Making the correct selection will be an important determining factor on how much value Financial Institutions derive from their AI enabled activities.
Horizontal AI tools are industry agnostic, they focus on doing generalist tasks well and they can often support that task across any industry. The sorts of tasks where a Financial Institution may decide to leverage a horizontal tool are things like meeting summarisation and note taking, generic email and calendar invite creation, customer chatbots, code creation and generic internal document search. These tasks don’t typically require a deep level of domain specific context to provide helpful support.
There are many tasks and mission critical activities, however, where these horizontal tools don’t have the context, knowledge and workflows to provide the support needed. This is where Vertical AI-SaaS tools come in. Vertical AI-SaaS tools deploy the required context, configuration and knowledge to provide domain specific, fit for purpose outputs. Such tasks include credit memo generation, investment memo generation, policy mapping or covenant testing. Today these are often completed manually and tomorrow they will be drafted or completed by AI and reviewed/edited/approved by front line teams.
These specialist activities are often complex and are far beyond the capabilities of Horizontal AI-SaaS offerings. They can also create a challenge for some Vertical AI-SaaS applications too given the high levels of prompt engineering and/or fine tuning required to deliver the relevance needed. High levels of prompt engineering hinder adoption and therefore limit success rates.
Covecta’s Configurable AI Platform addresses this complexity of specialist activities and tasks through its proprietary Vertical AI-SaaS capabilities. The need for fine tuning and prompting is removed through Covecta’s proprietary Configurator Engine, which delivers automated prompting at a user level. Thus delivering both the strong relevance needed and high user adoption.
Join the waitlist today to learn more.
