Home

[0:00] Every company that I talk to wants to use AI. Almost none of them are doing it correctly. I'm not talking about buying a ChatGPT license and having it help you write your emails. I'm talking about integrating an agent into your ecosystem and tools and putting that agent to work for you. Let's talk about what a typical product team looks like. You have your customer support reps, you have your product, and you have your engineers. And that whole team has to work together in order to handle the bugs that get reported by your users. What does that look like? Well, starts with the user encountering an issue. After that, they're going to open a ticket, open a chat, and report it to your customer support rep. They're going to gather the information and then attempt to recreate the bug and kick that over to product so that they can decide where that bug fits in the pri in the many priorities that the development team is already overwhelmed with. Let's now reframe that to what it would look like with agents being integrated into your team. The idea behind this is putting agents at the front line of your bug triage. So the customer support still interacts directly with the users. However, instead of kicking over bugs to product to be prioritized, we take a diversion and we send it to the agent. The agent has access to your communications. It has access to your codebase. It understands your software. It understands your goals. All of your context sits with that agent. And so when it receives a bug report, it is able to go in, look at the code, figure out what could be causing the bug, and in the best case scenario, it knows the fix, it applies the fix, and it gets that ready to deploy to the staging environment with the push of a button.

[1:50] Guess what? You just saved four different interruptions with the agent version. It's not really the bug fixes that cost, it's the interruptions that cost you. Most people think of AI as a tool, something that you open up in your browser, ask it a question, and get a response. That's version one. Version two is AI is something that you integrate into your team. It has access to your systems. It knows your project's context. It responds to events in real time. It's something that doesn't replace your developers, but it acts as a shield to keep them away from the distractions that prevent them from being able to do their best work.