Adam SandCEO of Roofing Business Partner

End of Coding? Introducing Low-Code Agent for Complex Tasks #chatgpt #ai #aiworld #coding #shorts

5 Sept 2025 · 1 min

Watch on YouTube

0:54

Adam Sand discusses a near-term shift toward agentic AI—systems composed of multiple specialized pretrained agents that coordinate work. He highlights the emergence of low-code/no-code platforms (built by notable technical founders) that let non-developers compose manager agents which invoke and delegate tasks to other agents, enabling workflows like passing checklists between agents and returning control when subtasks complete.

Chapters

The next stage of AI is really about agentic AI. This is about having multiple pre trained AIs working together. So the other problem is like making agentic AI has been really tough. You had to know how to code because you're saving context window by writing code with predefined instructions. So then there's low code and no code options coming to market. One of those concepts coming to market is agent AI which is built by the same technical co founder as HubSpot. You can now build an agent in a low code environment where one agent does things and then hands it to another agent. The next version of this that's really exciting with agent AI is that you can invoke another agent and come back. So you could hand the checklist over, say do this fourth checkbox and then have it come back. So this is like the next really cool level is that you can kind of have a manager agent that invokes a number of direct report agents.