Software interfaces have evolved from keyboard to visual point-and-click interfaces to touch. Conversational interfaces started some years ago as chatbots were introduced. With GenAI, the capability to create and understand text/audio/visual has leaped exponentially. As these conversational interfaces become capable of executing actions and performing tasks for human users (and other programmatic use cases), the concept of agents becomes a reality.
An AI agent is the natural evolution of a chatbot in that it can make some sort of decision, take action on behalf of a user, and adapt without the need for constant supervision. For example, OpenAI recently announced their latest model, o1 (https://openai.com/o1/), which spends more time reasoning before providing answers to a user, getting us closer to an agentic future.
Large enterprise software companies are also following suit by incorporating the capability of admins to create agents within their platforms. Citizen users can also create their own agents to help them perform tasks. In fact, AI agents can invoke other AI agents to perform micro tasks (like the concept of reusable software modules in software development).
AI agent examples from a select group of enterprise software platforms:
At Alten Capital we invest in technology services businesses. Please reach out to us to explore potential partnerships.