AI Assistant Approaches for Existing SaaS Applications

Future SaaS applications will have some kind of AI assistants as the primary way of interacting with the service. These AI assistants are more than support chatbots; they actively help users perform specific actions, troubleshoot issues, generate reports, and much more.

Here, I've described 3 examples from leading SaaS applications with great AI assistants.

Embedded AI Assistant

The embedded AI assistant integrates into the SaaS product deeply, remaining accessible across screens and workflows. It leverages specialized SaaS domain knowledge and user-specific context awareness, it tailors responses and actions to meet individual user needs. For example, Supabase Assistant supabase.com/blog/supab… knows SQL, and user’s schemas. These assistants excel at automating complex tasks such as data entry, querying, and setup, streamlining user workflows, but require most effort to convert a SaaS into AI powered experience.

AI Assistant as an Extension

AI assistants as extensions focus on specific functionalities within a SaaS platform, such as website creation or workflow definition. While powerful, their availability is often fragmented, limited to certain steps or starting points in the workflow. Think how Vercel added v0.dev as a front to its SaaS, but not present in the rest of the SaaS. Additionally, they lack deep context integration with backend systems, reducing their ability to provide, contextualized support across the SaaS product. Yet, an AI extension could be the long-waited unlock to your service, and adding it will not interfere with the rest of yoru SaaS.

3rd-Party AI Assistant SaaS

Third-party AI assistant platforms provide broader functionality, supporting tasks like website creation, full-stack app development, and workflow automation across multiple domains. These are new, AI-native services specilizing in one thing, do it best, and integrate with various SaaS platforms for deployment, hosting, or monitoring, enabling cross-platform utility. Think of bolt.new as the best full-stack app creation environment and Supabase or Netlify as deployment targets. Or specialized to cater to specific personas, such as beginner-friendly solutions like lovable.dev.

Do you run a SaaS? What is your AI strategy?

Share to help others find this note, and subscribe to generativeprogrammer.comwhere I share weekly AI news for Devs!