← Video Library · Case Studies
Training Devs to Build Without Coding—This Is the Future
Published June 29, 2025 · 86 views on YouTube
What is Howdy’s two-part training plan for engineers?
Howdy is building a two-part training module to prepare its engineering team for AI making traditional coding obsolete. The first part teaches engineers the skills to become AI engineers, such as building inference models. The second, trickier part trains them to build products without ever writing a single line of code.
Why is Howdy training engineers to build without code?
Co-founder Frank Licea says Howdy is bracing for a future where AI makes coding obsolete, but where a new class of experts emerges who know how to work with AI to build products without touching code. Howdy wants to be ahead of that shift by upskilling its own engineering team now.
What productivity goal is Howdy setting for its engineers?
Howdy’s goal for this year is for every engineer to be five times more productive than any other engineer on a typical company team. For 2026, the target rises to 10x more productive and 10x more efficient, driven by the combination of AI-engineering skills and no-code product building.
How does this compare to hiring AI engineers directly?
Licea notes that AI engineers with these skills currently command compensation packages ranging from roughly $500,000 to $1 million at companies like OpenAI, Apple, and Google. Howdy’s strategy is to upskill its own engineers into this tier rather than compete for that expensive external talent, while also making them the first generation of engineers who can build products without ever touching code.
Full video transcript
What are any big initiatives, innovations or expansions that people should be watching for? One of the things that we are trying to brace for is the AI revolution and coding becoming obsolete. We’re today already living in the future. So we have developed a ton — it’s like a two-part training module. In this two-part training module, the first part is how do we teach our engineers to build AI products — teaching them, for example, how to build inference models or anything. If you needed an AI engineer, you don’t have to hire a super expensive resource here in the United States. There are people all over the world. So that’s the first one: just giving them the skills necessary to be considered an AI engineer.
And then the second one is now how do we train them in building products without ever writing a single line of code. And that’s a trickier one, and that’s one that we’re more excited about. But it’s this idea that there will be a future where that happens, but there’s going to be experts — experts in terms of people being able to work with AI. And our goal is this year to have every single one of our engineers be five times more productive than any other engineer on a company team. We want people to walk away from a review and be like, “Oh my gosh, Frank, I’m telling you, your output is five times more than any other person that we’re working with,” which is a very big, crazy audacious goal. But next year, for 2026, it’s to be 10x more productive and 10x more efficient.
So those are the two things that we’re really doubling down on right now: upskilling our team to, one, be able to be these types of AI engineers that are currently making 500,000 to a million dollar compensation packages with OpenAI and Apple and Google and the like, and then also basically making them the first generation of engineers who know how to build code without ever touching it.