From Wheat to Chatbots: Understanding ML Engineers vs Applied AI Engineers

Let's imagine a farmer who needs to grow wheat in his field. He needs to take care of multiple things. For example:

1. The right season to grow wheat
2. Whether he should grow only wheat in his field or combine it with other crops
3. How much water should be given to crops over the season, and many more things

Similar to growing a crop, the engineers who build LLMs (Large Language Models) are ML Engineers. They need to think about:

1. Which algorithm to use to build the model
2. How big the model should be
3. How to get the relevant data to build the model, and many more things

From Wheat to Bread: LLM API Providers

After selling the crops to the manufacturer, they convert the wheat to bread, which they sell further to others such as cafes.

The cafes use the bread and sell different types of food made using the bread — such as grilled sandwiches, bread butter, vegetable sandwiches, and many more.

Similar to converting wheat to bread, companies like OpenAI and Anthropic provide their LLM APIs to engineers and other companies.

Applied AI Engineers: Building with LLM APIs

Those who develop different kinds of applications using these LLM APIs and their software knowledge are called Applied AI Engineers. Applications such as:

1. Chatbots
2. AI code reviewers
3. Lovable, and many more

Applied AI Engineers are similar to the cafes selling different kinds of food made using the bread.

Wheat-to-bread analogy comparing farmers, manufacturers, and cafes to ML engineers, LLM API providers, and Applied AI engineers

Which Path Is Right for You?

So you don't need to build a model yourself to become an AI Engineer. You can become one by building AI applications using LLM APIs.

If you want to build a model like GPT or Claude yourself, you will be becoming an ML engineer, where you need to have a strong foundation in mathematics and statistics.

I will be releasing a blog on a high-level overview of how LLMs work, so stay tuned for it.