Evolving Landscape of AI: Horizontal vs Vertical Models
Overview
82% of executives believe AI will reduce business costs
…But how do you go about it?
Before you pitch the idea of getting AI to your CEO, here are some considerations that may be useful.
- Which AI model suits your business requirements?
- What are the potential implications of each AI model?
- Are there other pros and cons?
Which AI model suits your business requirements?
Two distinct models are making waves in the realm of AI –
Horizontal AI models and Vertical AI models
Like their users in the SaaS world, these AI types differ in scope and focus, which impacts their effectiveness in the enterprise landscape.
As a Salesforce professional, understanding these differences is crucial in making informed decisions about the AI strategy that best suits your business requirements.
| Industry | Usage | AI Use Cases |
|---|---|---|
|
Manufacturing
|
40%
|
|
|
Banking & FinServ
|
37%
|
|
|
Government
|
26%
|
|
Source: Gartner
The Horizontal AI Model
The strategy revolves around a single AI software that grows its roots in different aspects of a business model, like OpenAI, Claude, Cohere, etc. A single vendor powers this approach and focuses on broad datasets like the Salesforce Data Cloud. The unified model offers cross-functional business capabilities.
It can provide integrated solutions that organizations with investments in Salesforce require. Moreover, it addresses concerns related to integration and security, paving the way for expansion into a diverse AI ecosystem.
Drawbacks of Horizontal AI:-
Extensive Training Required
Companies offering horizontal AI solutions provide tools that require extensive knowledge of using AI. Employees must be trained to use AI.
Rise of the Vertical AI Model
On the other hand, vertical AI solutions focus on incorporating AI into specific industries, functions, or processes. These industry-specific capabilities are powered by multiple vertical AI strategies. All functions are finely tuned for specific nuances and trained on vertical datasets.
These vertical AI models are interwoven across applications to address unique industry challenges and offer an enriched user experience.
Organizations can benefit from using a financial services-specific artificial intelligence model such as Bloomberg GPT along with Salesforce and GPTfy.
This lets businesses focus on delivering the core value without applying effort in the heavy lift required to train industry-specific machine learning models. The challenge of having anonymized data relevant to an industry is fairly significant.
Using vertical AI eliminates that challenge and lets organizations move quickly and benefit from the rise of artificial intelligence as a driver for digital transformation.
Drawbacks of Vertical AI:-
Versatility Limitations
Specialized AI systems excel in specific tasks but may struggle to adapt or generalize their knowledge.
Bias & Overfitting
Training AI systems on domain-specific data can lead to biases and an increased risk of overfitting. Safeguards are crucial to ensure fair and accurate outcomes.
Horizontal AI Layer and Data Integration
Both AI models rely on a horizontal AI layer that covers the entire enterprise and feeds into various business-specific applications. This data layer plays a critical role in training AI models. Identifying use cases drives initial efficiency and business growth.
Companies must be open to constantly changing their strategy as AI adoption makes progress. It is essential to consider the complexities of security, privacy, compliance, and ethical considerations in AI implementation.
Balancing Hype and Real-World Scenarios
AI hype continues to grip the business world. It’s crucial to follow a few steps to thrive:-
- Differentiating between realistic expectations and exaggerated claims.
- AI must be utilized for practical applications, such as automating basic processes to prove business value.
- Gradual scaling and deploying AI capabilities based on real-world results can lead to meaningful transformation.
Bring Gen AI In Your Salesforce
Why spend thousands of dollars when you can get a Free trial of AI in your Salesforce securely and in compliance with regulations?
Get NowWatch DemosStep into an AI-Enabled Future with GPTfy
Empower your workforce to thrive alongside generative AI through robust functions. Enhance efficiency, productivity, and CSAT to drive improvements in operations.
GPTfy is the first secure prompt engineering platform for Salesforce. It comes with dozens of pre-engineered prompts ready to go in Sales Cloud, Service Cloud, and Financial Services Cloud.
It automatically masks and de-masks data as it is sent out to your chosen LLM API service. Additionally, it lets you use all the fine granularity of Salesforce security. This lets you assign permission sets and other security settings to each prompt.
Organizations using Salesforce can benefit by starting with horizontal AI to focus on simplicity. Direct yet effective use cases such as case and opportunity summarization, hyper-personalized email generation, sentiment analysis, and others.
Once the AI capability in the organization is mature enough, it may be good to pivot into vertical AI. Great examples of this are fraud detection claims processing and other industry-specific AI LED automation.
However, there may be nuggets of ROI that determine choosing a horizontal, vertical, or a combination of both.
We are curious to hear your perspective in the comments section.
Saurabh Gupta
Saurabh is an Enterprise Architect and seasoned entrepreneur spearheading a Salesforce security and AI startup, with inventive contributions recognized by a patent.
Blogs you may find interesting
Salesforce + ChatGPT: Do more with less
In the last few months, We met a lot of sales and service folks. And what we heard from them shocked and saddened us
How is technology hurting your sales users?
66% of Sales rep says that they are drowning in tools, how they negatively impact your team’s productivity and success
Generative AI and Salesforce CRM: How to utilize it?
Generative AI is here and will explode and become part of every industry and revolutionize them. We will see things we have never seen before.