Dave Ebbelaar

Making an LLM API call is the most expensive and most dangerous operation in modern software development.

While incredibly powerful, you want to avoid it at all costs and only use it when absolutely necessary.

Most successful AI applications I've seen are built with simple, custom building blocks, not agent frameworks.

This is because most effective "AI agents" aren't actually that agentic at all.

They're mostly deterministic software with strategic LLM calls placed exactly where they add value.

The problem is that most frameworks push the "give an LLM some tools and let it figure everything out" approach.

But in reality, you don't want your LLM making every decision.

Now there are exceptions to this (e.g., Cursor)

But generally, you want it handling the one thing it's good at - reasoning with context - while your code handles everything else.

The solution is simpler than most frameworks make it seem.

Here's the approach that actually works:

1. Break down what you're actually building into fundamental components
2. Solve each problem with proper software engineering best practices
3. ONLY INCLUDE AN LLM STEP when it's impossible to solve with deterministic code

But when you do make that LLM call, it's all about context engineering.

To get a good answer back, you need the right context at the right time sent to the right model.

You need to pre-process all available information, prompts, and user input so the LLM can easily and reliably solve the problem.

This is the most fundamental skill in working with LLMs.

AI agents are simply workflows - directed acyclic graphs (DAGs) if you're being precise, or graphs if you include loops.

Most steps in these workflows/graphs should be regular code - not LLM calls.

Given all of this, you only need about seven core building blocks to solve almost any business problem with AI.

Take your big problem, break it down into smaller problems, then solve each one using these building blocks chained together - that's how you build effective AI agents.

P.S. I explain all of this in more detail in the video below.

4 months ago | [YT] | 60

Dave Ebbelaar

Hey everyone, I hope you've had a lovely Christmas!

As we're wrapping up the year, I wanted to share my plans for 2025 with you to grow Datalumina.
In this video, I'll go over the master plan that the team and I created.

Here are some things I'll go into:

- The traffic system that runs 24/7, driving our leads
- The SaaS product fueling our pivot to scalable growth
- Why we’re minimizing custom development to focus on consulting and productized services

If you’re curious about how to structure a business with synergy across products and services, this behind-the-scenes look is for you.

P.S. I tried a new style of video where I show you more behind the scenes of what's going on in my life. I plan to do this more often next year. I hope you like it!

11 months ago | [YT] | 12

Dave Ebbelaar

After two years of building with GenAI, here’s what I wish I’d had from day one...

This has been a long time in the making, and I’m excited to finally pull back the curtain on what we’ve been building behind the scenes:

The GenAI Launchpad — officially launched on Product Hunt today! 🎉

For the past two years, my team and I at Datalumina have been deeply involved in the world of AI, building solutions with large language models (LLMs) for clients across industries.

Each project taught us so much about what it takes to bring AI to life in practical, high-impact ways.

But there was one recurring challenge...

We spent way too much time setting up project structures, handling integrations, and putting out fires in the infrastructure — leaving less time for the real AI work, the work that brings ideas to life.

Not only was setup eating into our time, but we also found that the agent frameworks on the market were just too optimistic.

Real-world use cases are more complex and demand reliability and precision that many frameworks simply can’t deliver.

So, we got to work! 👷🏼‍♂️

And after two years of trial and error, working with every system and structure you can imagine, we built our own solution.

The GenAI Launchpad is the result of our journey — a project repository that streamlines everything from initial setup to deployment, ready to handle the demands of production at scale.

And the time savings? ⏳

We’ve calculated that it saves us over 50 hours per project on average, so we can dive right into the creative work that actually advances AI.

Today, we’re launching the GenAI Launchpad to share that time-saving power with you — our community of fellow AI enthusiasts and builders.

This is more than just a repository; it’s a battle-tested, engineer-approved blueprint that I wish I’d had when we started.

It’s here to help you skip the headaches, bypass the boilerplate, and focus on what matters: building innovative AI solutions for real-world problems.

If you’ve ever spent weeks fighting project setup, only to finally reach the real work, then you’ll understand why I’m so excited to share this.

👉🏻 Check it out on Product Hunt today:
www.producthunt.com/posts/genai-launchpad

If you’re curious to see what it’s all about or just want to support our work, I’d love for you to check out the page.

And if you’re already active on Product Hunt, it would mean a lot if you’d drop by to share your feedback or leave a comment!

1 year ago (edited) | [YT] | 80

Dave Ebbelaar

When I hit ‘record’ for the first time, I had no idea it would lead to this...

But it was one of the best decisions I've made.

It’s hard to believe, but I've just hit 100,000 subscribers on YouTube! And I want to share a bit of that journey with you.

My love for making videos started way back when I was just 10 years old. Me and my friends would grab whatever camera our parents had laying around and started creating all kinds of short movies.

By the time I was 16, I turned that passion into my first “business” as a freelance videographer.

But then, as I transitioned into full-time freelancing as a data scientist, I started to miss making videos.

And that’s how this YouTube channel was born, about 2.5 years ago.

It started with the idea of filling a gap I had noticed for a long time.

So many AI/ML tutorials out there were either overly complex, boring, or poorly produced.

I wanted to create practical, simple-to-follow tutorials that get straight to the point.

Now here we are, 100K strong.

I’m grateful for every single one of you who’s been part of this journey.

How long have you been following the channel? And what keeps you coming back to watch my videos?

I’d love to hear from you!

1 year ago | [YT] | 188