Gusto’s head of technology says hiring an army of specialists is the wrong approach to AI
As founders plan for an increasingly AI-centric future, Gusto co-founder and head of technology Edward Kim said that cutting existing teams and hiring a bunch of specially trained AI engineers is “the wrong way to go.”
The Problem with Traditional AI Development
Traditional AI development involves hiring specialized AI engineers who have extensive knowledge of machine learning, programming languages, and data science. This approach can be costly, as companies often need to pay high salaries to attract and retain tOP talent. Additionally, the traditional approach can be siloed, with AI engineers working in isolation from other teams, such as customer support and product development.
Gusto's Approach: Empowering Non-Technical Team Members
Gusto takes a different approach by empowering non-technical team members, particularly those in customer support, to write "recipes" that guide the AI assistant Gus. These recipes are essentially instructions in natural language that teach Gus how to handle specific customer scenarios.
The company's approach is based on the idea that non-technical team members have a deeper understanding of customer pain points and needs. They are closer to the customers and have a better understanding of the complexities of the business. By leveraging this expertise, Gusto can build AI applications that are more relevant and effective.
The Recipe-Based Approach
The recipe-based approach involves non-technical team members writing recipes that describe how Gus should handle specific customer scenarios. These recipes can be written in natural language, using everyday language that is easy to understand. The recipes can include information such as:
The recipes are then used to train Gus, which can use the information to make decisions and take actions that meet the customer's needs.
The Benefits of the Recipe-Based Approach
The recipe-based approach has several benefits, including:
Case Study: Eric Rodriguez's CoPilot Tool
One example of Gusto's approach is Eric Rodriguez's CoPilot tool, which was built by a customer support team member. Eric's tool is a natural language processing (NLP) application that can answer customer questions in natural language. The tool was built using a combination of NLP and machine learning algorithms, and was trained on a large dataset of customer support interactions.
The CoPilot tool was a success, and was adopted by the company's customer support team. The tool allowed the team to respond more quickly and effectively to customer inquiries, and improved the overall customer experience.
The Future of AI Development
Gusto's approach to building AI applications is likely to become more prevalent in the future. As AI continues to evolve, companies will need to find ways to leverage the expertise of non-technical team members, such as customer support and product development teams.
The recipe-based approach will likely become a key component of this, as companies seek to build more human-centered and customer-obsessed AI applications. By empowering non-technical team members to write recipes that guide AI assistants, companies can create more effective and efficient AI applications that meet the needs of customers.
Key Takeaways
Article