De SaaS a Agentes: el cambio que viene (y cómo aprovecharlo)

Summary

The video explains the transition from the traditional SaaS model to the 'Agents as a Service' model. Unlike conventional software where the user utilizes tools to perform tasks, here the product is an autonomous agent that executes the entire job for the client. The technical roadmap required to build these agents is detailed: understanding LLM reasoning loops, managing context through vector databases (RAG), integrating tools via the MCP standard, and establishing human oversight mechanisms to avoid errors and unforeseen costs.

Topics covered

• Difference between SaaS and Agents as a Service
• How agent reasoning works

+5 more topics (full content when you follow the channel)

Transcript excerpt

What you see on the screen is one of my SAS. Oh, it seems so, because underneath it 's an agent working alone, hidden behind a UI that pretends to be a normal application. It's called Combo Hunter and it monitors forums and networks to find people asking for exactly what you sell. Many have begun to call this Agents Assas Service, a new model growing in parallel with SAS. Instead of selling the customer a tool for him to do the job, you build an agent that does it for him. Your real product is the agent, not the application. And as building software becomes increasingly easier, simple apps become commodities. To differentiate yourself, you need to build products that do more for the customer. In this video I'm going to give you the real path to building agents and selling them as agents [music] as a service. Where to start, what problems, and how to level up your products? Let's…