.svg)
Machine Learning Engineer
Do you enjoy solving complex problems?
Are you curious about how artificial intelligence (AI) can predict trends and behaviors?
Would you like to be part of a team that builds intelligent systems, from self-driving cars to personal health assistants?
Dive into the world of AI to discover exciting projects you could work on as a Machine Learning Engineer.
‘Machine learning’ means getting computers to act without being programmed. This is the basis of Artificial Intelligence. Simply put, Machine Learning Engineers are the people who “teach” the machines by first feeding them data. They then create programs that teach machines to make predictions using that data.
You may be thinking that “engineer” sounds like a difficult job that requires years of study before you can earn any money. But ‘engineer’ is a word for people who design and manage systems, and there are lots of levels of engineering from simple to complex. As AI gets more and more widespread, the demand for Machine Learning Engineers will be huge, and there will be easier entry points to do this important work.
Let’s break it down and explain how you can get started quickly.
Want to try it out now?
What does a Machine Learning Engineer actually do? In this exercise, you’ll try out a version of what their work is like. You will generate some data, train the model, and test the machine learning to see if it is working.
Try It Out - Machine Learning Engineer
-You will need a camera on your phone, tablet or laptop
-You will need to be in a place where you can move around
Step - 1 : Set up your Machine
Navigate to this You should see this:
Click in ‘Class 1’ and change the name to “Y”. Then click on the webcam under “Y”.
It will ask permission to use your camera - click Yes.
You should see this:
Click on the settings button next to ‘record’. We need to make it easier for you to record your training data.
You should see this:
Click ‘hold to record’ to off. Add a 2 second delay. Make duration 3 seconds. Then click ‘save settings’.
Practice making the shape of a ‘Y’ using arms. Then click on ‘record’. Make sure the webcam can see the whole of your upper body and arms. Wait for it to take pictures of you.
Now go down to the box which says ‘Class 2’. Rename this to “M”. Your settings should have carried over into this box. Click ‘record’ and make an ‘M’ shape with your arms. Wait for it to take pictures of your ‘M’.
Now we need to add one more class. Neutral. Below ‘M’ click on ‘add class’:
Rename this new class to ‘neutral’. Click ‘record’ and leave your arms and face in a neutral position.
Step - 2 : Train Your Machine
Now the fun begins! Hit ‘train model’
Wait for the model to be trained. You need to keep your browser tab open.
Note: due to the amount of data being used this site can be a bit slow - be patient!
Once the model is trained, you try it out on the right. Make a ‘Y’ - does it recognise it? The percentages are the percentage probability that what it is labeling it as is correct.
What is going right? What is going wrong? How might you improve the model?
How does this machine learning task you just completed apply in the real world?
An example is self driving cars. Self driving cars use cameras to ‘view’ their surroundings. One of the things they need to predict is the future behavior of people that are near the roadside that the car will pass. By creating a model that uses the poses of people, then using a lot of training data, the AI in the car can predict if someone is about to step out in front of the car. This might slow or halt the car in response - saving someone’s life!
Another example is at airport security. The Transportation Security Administration (TSA) hasAI systems that are trained to recognize images and detect threats. After being trained on hundreds of thousands of images, AI can predict which objects in bags are suspicious. As your bag goes through the scanner, the machine tries to identify suspicious objects based on their shape and density. If it sees something suspicious, it flags the image to the agent for further inspection.
Can you think of other use cases?
Ready to go deeper?
To see some real-life and current machine learning projects check out the free site Kaggle. Kaggle is a community for data science and machine learning practitioners, hosting competitions to solve problems with machine learning. Check out the latest competitions here: https://www.kaggle.com/competitions
Start with the Titanic Machine Learning example. Your task is to use machine learning to create a model that predicts which passengers survived the Titanic. Watch the video first. If you decide you want to enter the competition, you can use this tutorial.
Understand the Business
Machine Learning Engineers work in many different environments, from cutting-edge tech startups to established corporations.
In the startup world, over 1,000 AI companies have been founded and funded in 2023! Most - if not all - of these companies need machine learning engineering. Some of these companies are developing ways AI can help farmers plant the right crops, or developing personal agents that help kids to learn math, or using models to reduce food waste in restaurants.
Most large organizations are also hiring machine learning engineers, especially in areas like financial services. Here you might work on developing new techniques for detecting fraud, or making it easier for customers to find the best way to cut their debt interest payments.
Here’s an example of how machine learning is being used in the airline industry.
The Job Outlook
The demand for Machine Learning Engineers has skyrocketed in recent years, making it one of the most sought-after careers in the tech industry. This trend is expected to continue as more industries adopt AI technologies.
As AI rapidly grows, there will be thousands of new AI engineers required. The Department of Labor forecasts that demand will increase by 36% per year
Salaries in this field are highly competitive, reflecting the high demand and specialized skill set required. In NYC a Machine Learning Engineer can earn around $160k a year.
Career Paths
As the demand for Machine Learning Engineers grows and the range of jobs broadens, there will be more entry points to enter the field. Many people get their start in Machine Learning through a Data Analyst path, although many other tech careers can be steps along a path to becoming a Machine Learning Engineer.
At an entry level, you might start out with tasks like Data Management to feed into AI programs, or work as an AI Tester who tests the performance of AI programs. In this role, you would first help train the AI program with training data. You’d then check the programs outputs by using testing data. You are identifying where the program is making the correct outputs compared to that data. If errors are occurring, you might then look for ways to better train the program to avoid those errors.
After gaining experience, you might move into more senior roles with broader responsibilities such as building the AI programs themselves.
It’s typical to get a start in the field through a Bachelor's or advanced degree. There are also free courses from Google, Microsoft and Facebook that can help you get started, which include industry recognised certifications For example, Google has a Machine Learning and Artificial Intelligence course. IBM Skillsbuild (free with sign-up) has an Artificial Intelligence course.
If this is a career you want to pursue, there is a path for you. Click on the videos below to hear how some Machine Learning Engineers got started.
Smitha Kolan is a self-taught MLE
Jordan Harrod has some insights
Skills to Pay the Bills
Machine Learning Engineering is a blend of technical and non-technical skills. Being a strong problem-solver is essential, along with a good grasp of math and statistics.
You’re learning basic programming skills, especially in a language like Python
You’re curious and willing to persevere through challenges.
You like to work collaboratively in teams
Skills you will Learn
These are some of the skills you will get good at as you continue on your path to a MLE:
You’ll be very familiar with data structures, algorithms and software design principles
You’ll learn about neural networks, data mining and predictive analytics.
You’ll learn to build and use your knowledge of coding programs like Python or C++
You’ll add to your mathematic knowledge - including linear algebra, calculus, probability and statistics
Review a Job Description
Job Descriptions are the way a company recruits and hires talent. You can learn a lot from a Job Description about the skills and qualifications you need to prepare for the job. Even if you are not ready to apply for a job now, learning to read the Job Description can help you prepare for what you need to do after high school.
Job Description
Company Name Withheld | is hiring a Junior Machine Learning Engineer.
Qualifications
Assist in Developing Machine Learning Systems:
Support the development of machine learning models under supervision.
Contribute to the implementation and maintenance of machine learning solutions.
Stay on top of new models that might help with implementation
Data Handling:
Assist in preprocessing and analyzing datasets.
Help in identifying data-related issues and resolving them.
Model Testing and Documentation:
Participate in model testing and provide feedback.
Document machine learning processes and results.
Collaborative Work:
Collaborate with senior engineers and data scientists.
Participate in team meetings and contribute ideas.
Learning and Growth:
Continuously learn and stay updated with machine learning trends.
Take on progressively challenging tasks as skills develop.
Desired Skills
Basic understanding of machine learning algorithms and principles.
Eagerness to learn and adapt to new technologies and methods.
Good problem-solving and analytical skills.
Basic familiarity with data processing tools and frameworks.
Ability to work in a team
Outstanding analytical and problem-solving skills
Effective communication skills.
Nice to have
SaaS (Software as a Service) knowledge or experience
Mobile app development experience
Experience with Google Analytics, Firebase, and Google Tag Manager
Interest in energy alternatives and combating climate change
