How to get into Artifical Intelligence Software Engineering — 2023

Greg Gompers
6 min readAug 27, 2022

This is a short list of information I wish I had when I was starting out.

Enjoy, and message me if you have any specific questions from here, Id love to help you out in your journey. Starting with no coding experience, and becoming a Machine Learning Engineer has been one of the best things Ive ever done


Definintion of words

  1. “Artificial Intelligence Algorithm” — lets break this down into each of the 3 words here.

“Artificial”— A human-made thing

“Intelligence”- The ability to learn things, and do things

“Algorithm” — A math equation

So, an “artificial intelligence algorithm”, is, a human-made, math equation, that is able to learn things, and do things. (useful things).

2. Artificial intelligence = Machine learning

These two terms mean same thing. People who are more familiar with the industry tend to prefer the term “Machine Learning”, and “ML” is the abbrievation.

3. “Data” —

Think of this word, being the same as… “measurements.” For example, if you wrote down the “measurements,” of the shoes you are wearing right now, you would probably write down things such as, their color, their size, their weight, their brand, their cost, your date of purchase, etc. After writing these things down on a piece of paper, you now have recorded some of the “data” about your shoes. Another way to say this, is that you now have collected a “dataset” for your shoes.

4. Machine Learning Algorithm — a more detailed explanation

This is a (long) math equation, that is able to read a specific “dataset” (about your shoes, for example) and use a combination of “statistics,” “calculus” and “linear algebra” to make a prediction on it.

Understanding the math inside of the equation is actually not that important. For example, do you need to understand the math going on inside of your cars engine, just to drive it? No, not really. So machine learning is the same way, 99% of the people just hop in the car and drive. That being said, the math going on in these equations is pretty cool if you’re into math, but if not, all good. Turn the key on the algorithm, and enjoy the ride.

AI is all about using a (long) math equation to making predictions on things that havent happened yet.

As another example — if you collected the “data” about all the houses on your street, such as, their number of bedrooms, the number of bathrooms, sqft, color, yard size, does it have a pool? Etc, then a machine learning algorithm would be able to predict the sale price of the house, just based on looking at the sale price of similar houses.

Now, yes, a human can also do this, but an ML algorithm can do this 1000X faster, and with a greater accuracy, that’s why ML is gaining so much popularity now

5. Deep Learning —

This is just the term used for the newest version of “machine learning algorithms.” You will see “machine learning” used as a more generic word, and when you see the word “deep learning (algorithm)”, just know this is a machine learning algorithm, with a slightly newer “architecture.”

Aka, a newer math equation that can do different things and/or is more efficient than the earlier, simpler, “machine learning algorithms.” For example, the only way to build Dalle-2, is by using a Deep Learning algorithm (a “deep learning” math equation).


The 3 steps that every AI software engineering project will go through

  1. Data Engineer

Ensures that all the data that is recorded is correct, organized, and stored in the most efficient way possible. 100k starting. “DE” = Data Engineer

2. Data Scientist

Takes all of the data organized by the Data Engineer, and puts it into a machine learning algorithm which is able to make predictions about this data, and/or make a decision based on this data. Or, in the case of Dalle-2, generate a completely new image, based on the “data” you have given it, (the text prompt you give it). 100k starting. “DS” = Data Scientist

3. Machine Learning Engineer

Takes the final version of the machine learning algorithm, built by the data scientist, and “scales” it, with cloud computing power. This is what allows Dalle-2 to generate an image in 2–4 seconds, because it’s code is running on 10 “ massive computers”, in 10 different locations across the globe. 100k starting. “MLE” = Machine Learning Engineer


Other job titles

  1. Python Engineer

This is is the first job you can get once you are familiar with python, its a “generalist” kind of job, it can be a bit quicker to get into, because all that is needed is knowledge of python, since only python is needed, it pays a little lower on avg, around 80–90k starting

2. Software Engineer

This is a very generic word, technically all of these roles above are considered types of “software engineers,” however, most people use the generic term, “software engineer” to describe a “javascript” software engineer, which is another way of saying a, “web development engineer,” and as a side note, “web developers” often are able to build apps on the iphone as well, because this iphone app code is very similar to “web dev” code. Also, People who specialize only in making mobile apps, are called “mobile app developers.”


Hype links that show what AI can do today, and where it is headed in the future:

  1. OpenAI Five Beats World Champion DOTA2 Team 2–0! 🤖

2. OpenAI GPT-3 — Good At Almost Everything! 🤖

3. — — 3min video on this page, the future of graphic design


Steps to becoming any of the 3 AI related roles above

  1. Learn Python

All of the AI related roles use python. Libraries such as “Pandas”, “Numpy”, and “Sklearn” are the most often used

“Libraries”, are like “expansions” that are built on top of python, that make it easier to do things, in less lines of code

“Sklearn” is the most common python library that makes it easier to create the older, simpler ML Algorithms. Along with this, “pytorch”, and “tensorflow” are the newer “libraries” that will allow you to create the newer kinds of algorithms, needed to do ChatGPT and image generation types of tasks.

“Datacamp” is a decent platform for this, also a friend of mine is using a platform called “Pirple”, may want to check this one out as well

Best option: BrainStation Python Bootcamp — 5 weeks — This one looks very high quality, always remember, the more high quality and structured the course is,and the more you have social interactions with classmates and the instructor, the faster and better you will learn it. Expect these setups to cost more, but it ends up being 100% worth it because of the increase in quality and learning speed. (Paying 3,000 to get hired with 100k salary, 2 months sooner, ends up being a net gain of +$9,000)

2. Learn the basics of the role you find most exciting to you (data engineer/ data scientist/ machine learning engineer/ software engineer)

I personally started with data science, and did this for quite some time, but at a certain point I realized I really liked Machine Learning Engineering the most, because it involves doing a little bit of all of the roles, so I found that this role ends up having the greatest understanding of the whole process which is something I found very useful. Machine Learning Engineer Nano-Degree

3. Take a 4–6 month BootCamp for the role

Ideally you want this to be a program that is highly social, so that you are able to meetup with classmates 1–3 times a week, and be surrounded by people already in the industry, which can answer all of the questions you will have along the way. The right program will also work with you 1:1 to build your resume, your linkedin profile, do practice interviews with you, and ensure you have everything you need to get started working in the field. If you do all the coursework, and actively ask all your questions, this guarantees your entry into the industry.

1. UC San Diego (SpringBoot) — Machine Learning Engineer Bootcamp — this is one of the best ones I can see for the MLE role

2. FourthBrain — Machine Learning Engineer Bootcamp — another option, newer company, and very socially collaborative, I did the level 2 version of this bootcamp with this company, very solid.

When I was first deciding which bootcamp to take, my decision was between the 2 bootcamps above


Again, feel free to message me for any questions you have from here, I will update this post as I get more questions