Last week, we hosted a webinar to talk about something on everyone's mind: how to actually build and ship AI products without the process turning into a security or operational nightmare. We polled the room and found that while a few teams are scaling, the vast majority are still in the experimentation phase.

If you missed the live session, here is the breakdown of what we discussed, from the changing role of engineers to our internal framework for moving fast.

How is AI changing the role of software engineers and product leaders?

Lucas Ward, founder of Auditless, joined us to share a bit of history. He pointed out that back in the days of COBOL, programmers were often the ones who understood the business better than anyone else. Somewhere along the line, the industry drifted into "architect" roles that felt divorced from how a company actually makes money.

AI is changing that. Lucas argued that we are on a "leverage increasing equation" similar to the steam engine or the industrial revolution. With AI handling more of the technical heavy lifting, developers and product leaders have to get closer to the actual business problems. At his company, even non-technical staff are using tools like Claude and VS Code to do meaningful work.

Can non-engineers build real products using AI?

Our second guest speaker, Jeremiah Johnson from Gladiator Tennis, shared a story that likely resonated with every non-coder in the audience. Tasked with building a custom administration system for 2,000 tennis players, he didn’t hire a massive dev team. Instead, he experimented with AI tools like Cursor and Claude.

In just two months, he built an MVP that runs their entire business. Jeremiah reminded us what it used to take to "become a master" in a technology a while back. In college, he had to read a 600-page PHP book to understand the language. Now, AI allows him to be a "master of outcomes" rather than a master of syntax. He isn't looking at code daily. Instead, he's collaborating with AI to solve specific problems for his users.

The Jetpacks! Framework: How do companies move AI from prototypes to real systems?

Our team walked through the tool-agnostic framework we use to take ideas into production:

• Problem Discovery: We use AI to help with market research and to stress-test whether a problem is actually worth solving.
• The PRD (Product Requirement Document): This is the most important step. We use AI to draft technical and functional requirements, including security and compliance constraints, before a single line of code is written.
• The Build: We don't code manually. We provide the PRD to AI agents and let them execute the stories.
• Launch and Iterate: We launch fast to find mistakes while they are still cheap to fix.
• The Human-in-the-Loop: A common fear is that AI will just "take over," but our framework relies on a clear division of labor.
• AI is the Orchestrator: It handles the manual work, like crafting stories or generating boilerplate code.
• The Human is the Strategist: People own the strategic vision and the business goals. You can delegate the "doing," but you can't delegate the "deciding".

What does it really take for a company to be ready for AI?

We ended the session with a reminder from Jetpacks!’s CEO, Gaston: AI readiness isn't a tools problem, it’s about people and process. The teams that succeed aren't necessarily using the most expensive models. They are the ones where people feel safe experimenting and where processes are lightweight enough to allow for failure.

If you want to move from hype to actual architecture, start with your data foundation. Once your data is secure and accessible, you can build the "agentic workflows" that actually move the needle for your business.

If your team is stuck in pilots, demos, or internal experiments, the problem is not the tools. It is the framework. Jetpacks! helps teams design, build, and launch AI products with speed and control.

Let’s talk about what shipping AI looks like for your business.

‍