is the brand new “hot” role within the tech scene, and lots of individuals are eager to land this job.
I see so many posts online saying how you’ll be able to turn out to be an AI engineer in a couple of months.
Let me be clear: anyone telling you you can turn out to be an AI engineer in six months is selling you a dream.
The fact is that it’ll take longer, but that doesn’t mean you’ll be able to’t attempt to fast-track the method.
Should you’re recent here, I’m Egor. I work as a machine learning engineer and am also a profession coach for people breaking into data, AI, and machine learning.
I’ve seen firsthand what works and what’s only a waste of time.
Let’s get into it!
Let’s make clear exactly what an AI engineer is, as there’s a whole lot of confusion online.
I even have a separate article explaining the important thing differences, but in a nutshell an AI engineer is a software engineer who specialises within the use and integration of foundational GenAI models corresponding to Claude, GPT, BERT, and others.
They don’t “construct” these models from scratch like an information scientist or machine learning engineer; somewhat, they use them to serve a particular purpose.
For instance, they could embed a chatbot on a shopping website to assist customers find what they’re searching for more quickly, or add a coding assistant in an IDE, like Cursor.
As AI engineers are specialised software engineers, they should know the basic practises of software engineering and have a robust knowledge of AI systems.
This skillset is rare but in high demand nowadays attributable to the hype around AI. So naturally, the salary of AI engineers could be very high and lots of corporations pay around $200k–$300k, in response to levels.fyi.
As you’ll be able to see, it’s a fairly attractive profession with a whole lot of growth potential. Let’s now go over exactly how you’ll be able to turn out to be one.
One unlucky reality is that it is incredibly difficult to interrupt into AI engineering with zero previous experience.
It’s because the career requires sufficient expertise across data, machine learning, software engineering, and, naturally, AI.
Subsequently, you will have to turn out to be an information scientist or a software engineer for at the very least one yr before considering of pivoting to becoming an AI engineer.
Whether you turn out to be an information scientist or a software engineer is as much as you and your background.
Nonetheless, I personally recommend starting as a software engineer first, because it’s more closely related to the AI engineering role.
You furthermore may don’t must take my word for it; Greg Brockman (OpenAI CTO) also agrees that it’s higher to be a software engineer first after which improve your AI/ML knowledge.
As a software engineer, you need to make an effort to learn the tools and technologies which can be needed to turn out to be an AI engineer, these include:
- Python — The entire AI/ML ecosystem is in-built Python, so you need to have the ability to jot down solid production code on this language.
- SQL — AI revolves around data, and SQL is the language of Data.
- Software Development Tools — Have to know things like git for version control, zsh/bash basics and understanding find out how to create and use APIs.
- System Design Technologies — The AI system you’ll eventually construct might want to scale, and you’ll likely deploy it on cloud platforms like AWS, Azure, GCP using tools like Docker and Kubernetes.
Resources
Timeline
The timeline is dependent upon how long it takes you to land a software engineering or data science job.
Being practical, if you may have a STEM background with some solid knowledge, and you actually apply yourself, you’ll be able to land jobs in these roles in about 6 months.
You must then stay on this role for a few yr before attempting to make the switch to AI engineering to make sure you may have covered your basics.
There are various guides online on find out how to break into software engineering, and I even have several roadmaps for becoming an information scientist you can also take a look at.
Alongside your full-time work as a software engineer, you will have to up-skill yourself on the fundamentals of AI/ML to make sure you make quick progress in your journey.
You definitely don’t have to have a PhD in Maths level of understanding, as you won’t construct these models from scratch, nevertheless it gives you background details to dive deeper into more advanced topics at a later date.
These are the things you need to know:
- Maths Fundamentals — A solid overview of statistics, probability, linear algebra and calculus will allow you to understand what’s happening under the hood.
- Supervised Learning — Know the way the fundamental algorithms like linear regression, decision trees and support vector machines work.
- Unsupervised Learning — Know the way the fundamental algorithms like K-Means and K-Nearest-Neighbour work.
- Neural Networks — These are the backbone of LLMs, and having understanding of topics like backpropagation, vanishing gradients and activation functions will permit you to debug AI models quicker in the long run.
- Basics of LLMs — Though you won’t be constructing LLMs from scratch, you will probably be working with them daily, so it’s good to have some knowledge about how they operate. You must study areas corresponding to transformers, autoencoders, tokenisation, and embeddings.
Resources
Timeline
Learning the basics will rely upon exactly how long you study while working as an information scientist/software engineer.
The advice is to integrate these concepts into your every day work as much as possible.
If I were studying all of this outside of working hours, I’d anticipate it could take 3–6 months should you apply yourself.
At this point, it’s time to dive deeper into the precise concepts and concepts you will probably be using as an AI engineer in the true world.
This field is evolving rapidly, and each month there’s a brand new “thing” to learn. I’ll list the timeless fundamentals here as they’re by far crucial.
- AI APIs — Services like OpenAI’s API allow you to integrate powerful models without having to construct them yourself. That is the fastest technique to start constructing real applications with AI capabilities.
- Prompt Engineering — Learning find out how to effectively communicate with AI models is a vital skill. Well-crafted prompts can dramatically improve model outputs and are essential for getting consistent results.
- Retrieval Augmented Generation (RAG) — Understand find out how to connect with LLMs to external databases like Pinecone and use related information to enhance the accuracy of the AI model’s responses.
- Model Context Protocol (MCP) — The standardised technique to connect your AI models to external applications like files, servers and other apps.
- LangChain — That is the perfect package for working with AI models in Python. It provides all of the architecture you’ll want to construct and connect LLMs seamlessly.
- Positive-Tuning — Understand find out how to improve the performance of an AI model by training it on specific data so it is healthier at responding and giving outputs for a certain use case.
Resources
Timeline
Learning these concepts will take barely less time than learning AI/ML fundamentals, as there’s less material to cover.
I’d anticipate it could take about 2–3 months to learn every part to standard.
There’s a whole lot of confusion of what projects you need to construct to be able to get a job in AI engineering
To place it simply, the perfect projects are ones which can be intrinsically motivating for you and in addition profit some form of end user or client.
Listed here are the high-level steps:
- Idea — Brainstorm ideas and topics which can be personal to you and an issue you would like to solve. This could come from your individual thoughts and research; don’t look online or ask people like me for project ideas. Anything I offer you will immediately be a foul project for you.
- Data — Find novel and exciting data using public APIs, government web sites, web-scraping, etc. You must replicate the messy data you’ll encounter in the true world.
- Deploy — You could showcase your ability to deploy AI systems end-to-end. It will include data storage, data cleansing, model connection then some integration on the front end through an API and even an online app. You could match the work you will probably be doing as a full-time AI engineer as closely as possible.
- Document — Nobody will find out about your project should you don’t tell people about it. Do a LinkedIn post, write a blog article and add it to your portfolio. Be sure that your project has a transparent, well-organised README on GitHub so people can test it for themselves. Share your work as much as possible, as it’ll increase your probabilities of being seen by potential employers.
Timeline
Creating good projects and constructing a solid portfolio will take time. Ideally, you need to construct two top-tier projects should take you about 3 months in total. This assumes you’ll be able to dedicate 1 hour per day to constructing these.
This might be a complete post in itself, but let me offer you the high level 80/20 of what you need to do:
Resume
To write down an ideal resume, be sure that every part is specifically about AI engineering:
- Have your technical skills right at the highest with relevant tools and technologies for AI engineering roles.
- Make your projects clearly visible with metrics, figures, and, particularly, the financial impact.
- Keep it easy: neutral colors, single column, easy-to-read fonts, and only a page long.
- List your relevant experience as either a software engineer or an information scientist.
I even have a full article on find out how to make an ideal resume you can take a look at below, in addition to a ready-made template you need to use.
Make your LinkedIn profile obvious that you just are going for AI engineering roles:
- Your headline should contain “AI Engineer”, no “aspiring” please. For instance, who would wish to hire an “aspiring” dentist?
- Include keywords throughout your “About me” and “Experience” sections, but add them organically and don’t write paragraphs.
- Make your profile aesthetic with a transparent photo and a nice-looking banner. This makes an even bigger difference than you’re thinking that.
Referrals & Networking
Most individuals think they should construct a great deal of projects and take countless courses to face out and get a job.
Referrals are the golden ticket for any tech job.
Based on a study, referrals account for 7% of applications but 40% of all hires. Should you’re referred, you’re almost 6x more prone to get your dream job.
That leverage is crazy.
The way in which you get a referral is definitely fairly easy, and all it requires is a few confidence in your part.
- Find corporations hiring for AI engineers or corporations you’d prefer to work for.
- Browse their employees on LinkedIn and find someone just like you. This might be someone with the identical university and background, ideally an AI engineer as well.
- Connect and send them a DM containing something you liked about their profile, journey or anything personal. Never ask for a referral in the primary message.
- Chat to them and ask them questions on their work, projects and anything cool they’re doing.
- After a couple of messages, that’s while you ask for a referral or any feedback in your resume.
The method is so easy, the issue is individuals are just too scared to do it.
Nonetheless, I even have never had a foul experience, because you mostly lead with a compliment or an opener about them.
People love talking about themselves, and all you’ll want to do is come across as friendly and show that you just are concerned with them.
Timeline
Getting a job can vary lots, and it may well also come right down to luck sometimes. Nonetheless, by actually going after referrals and avoiding distractions from projects and courses, this could take 6 months.
So, to turn out to be an AI engineer, it’ll take you, optimistically, about 2 years, but you furthermore may have to land a job as a software engineer or data scientist first.
This will appear to be an extended time, but these roles are highly expert and pay ridiculous salaries. You may’t expect to do a few courses and walk straight into them.
If after reading this text, you really need to turn out to be an AI engineer, that’s great!
Nonetheless, like I just mentioned, you’ll want to turn out to be an information scientist first. Fortunately, in certainly one of my previous articles, I wrote precisely the steps I’d follow if I were to turn out to be an information scientist again.
I’ll see you there!
Join my free newsletter where I share weekly suggestions, insights, and advice from my experience as a practising data scientist and machine learning engineer. Plus, as a subscriber, you’ll get my FREE Resume Template!
