Becoming an AI Engineer: The Fast Track to a Future Career

What is an AI Engineer?

An artificial intelligence engineer is distinguished by their ability to use and integrate advanced AI models such as Claude, GPT, and BERT into practical applications. Unlike data scientists or machine learning engineers who develop these models from scratch, the AI engineer focuses on their practical application. For example, they might set up a chatbot on an e-commerce site to enhance user experience or integrate a coding assistant into an integrated development environment like Cursor.

This role requires mastery of fundamental software engineering practices and a deep understanding of AI systems. The scarcity of this profile in the market, combined with growing demand, results in high salaries, often ranging from $200,000 to $300,000, according to data from levels.fyi.

Step 1: Entry Point

Starting a career in AI engineering without prior experience is a significant challenge. This career requires expertise in data, machine learning, software engineering, and, of course, AI. Therefore, it is advisable to begin as a data scientist or software engineer for at least a year before specializing in AI.

• It is particularly recommended to start in the role of software engineer, as it is more directly related to AI engineering.
• Greg Brockman, CTO of OpenAI, also supports this approach.

As a software engineer, it is essential to master certain tools and technologies to transition into AI engineering, including:

• Python: An essential language in the AI/ML ecosystem, it is crucial to know how to write robust production code.
• SQL: Since AI relies on data, SQL is indispensable for manipulating it.
• Software Development Tools: Mastery of git for version control, basic knowledge of zsh/bash, and the creation/usage of APIs is necessary.
• System Design Technologies: AI systems must be scalable, often deployed on cloud platforms like AWS, Azure, or GCP, using Docker and Kubernetes.

The time required to secure a job as a software engineer or data scientist varies depending on your background. With a solid STEM background and a serious commitment, it is possible to achieve this in about six months. It is advisable to stay in this role for a year before moving towards AI engineering.

Step 2: Fundamentals of AI/ML

In parallel with your full-time job as a software engineer, it is crucial to strengthen your knowledge of AI/ML to advance quickly in your career. A PhD in mathematics is not necessary, but a basic understanding is essential to tackle more complex topics later on.

Here are the key concepts to master:

• Mathematics Fundamentals: A good understanding of statistics, probability, linear algebra, and calculus is indispensable.
• Supervised Learning: Familiarize yourself with basic algorithms such as linear regression, decision trees, and support vector machines.
• Unsupervised Learning: Understanding basic algorithms like K-Means and K-Nearest Neighbour is crucial.
• Neural Networks: Master concepts such as backpropagation, vanishing gradients, and activation functions.
• Basics of LLMs: While you won't be building these models from scratch, it is useful to understand how they work.

The time required to grasp these fundamentals depends on your commitment outside of work hours. With regular study, this could take between three to six months.

Step 3: AI Engineering Concepts

It is time to focus on the specific concepts you will use as an AI engineer in a professional context. This field is rapidly evolving, with new knowledge to acquire each month. Here are the timeless fundamentals to know:

• AI APIs: Services like the OpenAI API allow you to integrate powerful models without building them yourself.
• Prompt Engineering: Knowing how to interact effectively with AI models is an essential skill.
• Retrieval-Augmented Generation (RAG): Understanding how to connect LLMs to external databases to improve response accuracy.
• Model Context Protocol (MCP): The standardized method for connecting your AI models to external applications.
• LangChain: The best package for working with AI models in Python.
• Fine-tuning: Understanding how to improve an AI model's performance by training it on specific data.

Learning these concepts will take slightly less time than the fundamentals of AI/ML, approximately two to three months.

Step 4: Projects

There is some confusion regarding the projects to undertake to secure a job in AI engineering. In summary, the best projects are those that you are passionate about and that provide added value to a user or client.

Here are the steps to follow:

• Idea: Think of ideas and topics that interest you and a problem you want to solve.
• Data: Find relevant data through public APIs, government websites, or web scraping.
• Deployment: Demonstrate your ability to deploy end-to-end AI systems.
• Documentation: Share your project on LinkedIn, write a blog post, and add it to your portfolio.

Creating quality projects and building a solid portfolio will take time. Ideally, you should complete two high-level projects in about three months, dedicating an hour a day to them.

Step 5: Getting a Job

While this could be the subject of an entire article, here are the key points for writing an excellent resume.