Right now we have three sets of Gauntlet AI spouses on stage (both spouses attended/graduated from Gauntlet). I don't know what it says about Gauntlet that people later encourage their spouse to come, but it makes me really happy. nitter.cf/i/broadcasts/1yxBeMDkP…

We're making direct enrollment in our AI for Engineering course a lot easier. Price will vary from time to time (lowest in December); any seat reserved by payment good any time in next 6 mos. Cohorts start monthly, everything flexible and designed to be outside of work hours.

We've changed how our AI for Developers training works! Yes, we're teaching engineers how to wield AI. Architecture, pipelines, fine tuning, agents, sytems, etc. BUT NOW: Every engineer is going to build a production-ready capstone project under the watchful eyes of our team.

It's ONBOARDING DAY! Ramping up new cohort of our AI for Developer Productivity course that starts tomorrow. Insanely rewarding to watch folks dramatically speed up doing the thing they are very good at. May every company double eng. productivity 🙏🤲

Last chance to get your engineering team into our "AI for Developer Productivity" cohort that starts July 23. We train your team on everything necessary to leverage AI for max productivity of a software engineering team. Can build initial stuff for you too. DMs open.

AI in software engineering is simultaneously way overhyped and way underhyped. What’s possible using AI goes far beyond using simple code autocomplete tools copy/pasting from StackOverflow. Yet we're very, very far from "We don't need software engineers."

Build vs. buy in AI right now has a clear answer for most companies: LLMs? Buy (or use open source). Completed AI product? Build. Why? AI models _need_ to have the context of your data and your IP. Completed AI products are bad at that today, without exception.

A prototype from a BloomTech AI student: This multi-agentic system writes React components in their entirety (within the context of the company code base.) Generates fully-formed React/JavaScript + CSS + unit tests + Documentation, completely automatically.

Manually writing code is like calculus. Using AI to code is like statistics. Manually coding you are spending huge amounts of human time to custom craft the perfect code. It's expensive because human time is expensive. A smart AI model may write code that that will only be right 50% of the time. But, importantly, you can perform that process nearly instantly a huge number of times, until the probability of success approaches 100%. And it's cheap because relative to human hours compute is very cheap. Computers are machines that make doing things large numbers of times insanely cheap. It makes sense that we use them this way.

We sent a team out to study the software engineers who are using AI to increase their productivity 2-5x. I was blown away: we saw individual engineers comfortably doing what used to require entire teams.

It’s been WILD seeing how much companies can accelerate as we train engineers to use AI correctly. Realizing it may make more sense for us to come in, set up all of the AI infrastructure, set up the initial agents, and then train the *entire* engineering org on how to use it.

So far from the ~100 companies I've talked to, about 1% are close to seeing what the impact of AI on an engineering team can be. The 1% are cracked and would blow your mind. Making AI work well requires architecture/data piping/model fine tuning that virtually no one is doing.

Case study from BloomTech’s AI for Developer Productivity course: A student built a multi-agentic system for his company that automatically: * builds react components * writes unit tests * writes documentation. He estimates it has saved his team 30-50 developer hours in the past week. And the craziest part? We haven’t even fine-tuned the models yet! (That starts tonight.)

If you’re a software engineer who has only ever used GPT/Copilot/Cursor off the shelf just know you’ve only seen the tip of the iceberg of what AI can do for software engineers

♥️

Cool student project from this: An AI agent that looks at code, commit messages, and context from the project management system. Then automatically writes out a detailed, human-friendly description of everything that changed and what that means for customers.

Check it out: An entire lesson from BloomTech's AI for Developer Productivity course! Fundamentals of RAG (Retrieval-Augmented-Generation). How we enhance accuracy and reliability of generative AI models. This is the foundation we build on to give AI important context.