Starting something new can feel overwhelming—especially in AI, where the sheer number of resources makes it hard to know where to begin.

In this post I’m going to cover:

1. My AI Engineering Roadmap
2. The courses I’ve chosen to take and why
3. Topics I’ve covered so far and where I can see real world applications

I’m going to be 100% honest about the courses I’m doing and how I’m finding them so that hopefully you can make a decision on whether they’re right for you.

AI Engineering Roadmap

So just to make it very clear – I am starting from scratch; but I do have an engineering background from my university degree and I have done some beginner coding lessons on Codecademy from when I thought I wanted to move into software engineering. 

My plan has three very simplified steps: 

1. Learn the theory (6months)
2. Apply knowledge to real world problems for experience (6months)
3. Apply experience to my work

Whether the timeline I have set out above is achievable is unclear to me at this time. It seems reasonable enough – although all of this has to be done outside of my day job. I am planning to dedicate an entire post to the challenges of self-learning whilst working full-time so please subscribe if you would like to see that when it comes out! 

Choosing my Learning Platform

If you go online and search for ‘AI Engineering learning materials’ there’s an incredible amount of choice ranging from free YouTube videos to very expensive university accredited qualifications. There are pros and cons for all of the materials spanning across that range, but here are the key things I was looking for: 

1. Structure – I wanted to be able to see a roadmap of all the different modules I was going to learn and the order in which I was going to take them. 
2. Projects – I wanted there to be real world projects included in the course so that I could put my newly learned skills to the test. 
3. Certification – this might not be so important, but I wanted something that I would be able to put on my CV and share with my LinkedIn network to show people the work I’d been doing. 
4. Affordability – Unfortunately, I don’t have lots of spare cash available to spend on the learning platform, so it was important that whatever I chose I could afford. 

My Chosen Learning Platform

I landed on Coursera as my learning platform as it meets all of my criteria providing a structured learning programme with projects built into it and a certificate at the end – all for £37/month. 

If you’ve had any experience with Coursera or have used other learning platforms then please let me know by posting a comment!

The programme I chose was the ‘IBM AI Engineering Professional Certificate‘ – when I looked through the modules it contained all the key words that I’d seen in job applications such as Keras, TensorFlow and PyTorch to name a few (at the time of writing I have no idea what they mean either).

Before I dived into the course, I did look to see if there were any course pre-requisites to ensure that I wasn’t completely lost when starting. Sure enough it suggested two:

1. IBM Data Science Professional Certificate
2. IBM Applied AI Professional Certificate

Whether you actually need to complete these or it’s just a way of getting more money out of you remains to be seen, but I decided I wanted to do things properly and would complete the two pre-requisite courses before starting the AI Engineering course. 

I’m planning to complete a course every 2 months – I started in Feb’25 so let me tell you about where I am so far. 

Topics I’ve Covered So Far

The ‘IBM Data Science Professional Certificate’ has 12 modules: 

1. What is Data Science? ✅ 
2. Tools for Data Science ✅ 
3. Data Science Methodology ✅ 
4. Python for Data Science, AI & Development ✅ 
5. Python for Data Science ✅ 
6. Databases and SQL for Data Science with Python ✅ 
7. Data Analysis with Python  ✍🏻 (In-Progress)
8. Data Visualisation with Python
9. Machine Learning with Python
10. Applied Science Capstone Project
11. Generative AI: Elevate Your Data Science Career
12. Data Scientist Career Guide and Interview Preparation 

Eagle-eyed readers will have noticed that I have only completed 6 modules and there is only 2 weeks left in March so I am behind on my plan… I am more than aware – but studying is hard and as previously mentioned, that is a topic for a future post! 

What is Data Science? & Tools for Data Science

The first two modules were very much introductory, which was great for me! The majority of the lessons were mainly videos but I found them all quite interesting and there are lots of quizzes throughout to make sure you’ve been paying attention.

One use case that I found really interesting was how Data Science was used to understand why there were spikes in customer complaints regarding buses provided by the public transport network in Canada (I think it was Canada). 

It turned out it was linked to the weather – whenever there was heavy rain there would be a spike in complaints. It helped the agency realise that it wasn’t the service they were providing that was really causing the complaints, it was an external factor. Still, the information meant that they could explore additional precautions to add to their service when there was adverse weather – it really shows how data provides powerful insights that can lead to valuable action. 

Data Science Methodology

My brain likes to work in logical steps and so I really enjoyed this module as it outlines a series of steps that need to be applied to data science problems.

There are 5 steps outlined in the Data Science Methodology: 

1. Business Understanding – spending time really understanding the problem and business outcomes that need to be addressed before diving into a project. 
2. Data Collection & Preparation – does what it says on the tin; if you haven’t got the right data in the right format, it’s going to be very difficult to come up with any meaningful insights. 
3. Data Exploration & Analysis – this is where it starts to get interesting for me. Diving into the detail of the data using statistics and visualisation techniques (applying all the Python skills I’m going to learn) to find meaningful insights, patterns and relationships. 
4. Model Building – things definitely get more complicated here but this is where you would apply tailored algorithms and models to the problem and then train and evaluate the models using the data you have collected to see how accurate you are. 
5. Model Deployment & Monitoring – deploying your model into the real world and applying it to new data. For example, the model you have built could be used in the healthcare industry to determine the likelihood a person has of developing a certain disease based on their blood test results and lifestyle choices.

Again, this module was mainly video based with lots of quizzes to test your learning but what was nice was they used the same Case Study throughout to show you how a problem can be worked through from start to finish. 

Python for Data Science, AI & Development and Python for Data Science

I really enjoyed the two Python modules, there is lots of content and topics covered but Coursera is good as it introduces you to all of the relevant tools used for best practice such as Jupyter notebooks, GitHub and also gives your free access to several IBM tools. 

The learning method is mainly lab based for these modules, which is good as it gets you familiar with setting up Jupyter notebooks, importing the required libraries and using specific functions. 

I won’t go into lots of detail as I appreciate that this post is already long (if you’ve read this far THANK YOU) but one thing I will say is that if you wanted to, you can easily skip through the labs without actually learning too much as a lot of the code is written for you, you just have to run it. What is good is the ‘Final Exams’ they have for each module – these do not give you any pointers and really make you think about what your doing. I have definitely found these to be the most challenging and rewarding exercises and I have felt a real sense of achievement when I have solved the problem. 

Databases and SQL for Data Science with Python

This module took me a very long time to complete – not because I found it difficult or boring, but just because I had other stuff going on that meant I was very busy. I was beating myself up about it for a little while but then I realised if things take longer than planned, that’s ok! In the words of Dory from Finding Nemo “Just Keep Swimming”. 

When I actually found the time to work through this module I found it really interesting and massively enjoyed it. SQL & Databases is not something I’ve had any experience with before but as previously mentioned, my brain thinks in logical steps and so databases containing tables with rows and columns is something that I understand, and seeing how SQL could be used to manipulate the data across multiple tables and databases was fun! 

Final Thoughts

On reflection, one of things I am enjoying so much about learning again, is how it makes you think differently. 

For example, when learning how to scrape data from the web using Python, I was thinking how scraping the data recorded in customer complaints could be immensely valuable for the company I currently work for, providing valuable insights into deep detractors and providing focus areas for action. Maybe I’ll get the confidence to take it to them when I’m a bit more well versed in the technical details? 

Will any of this help me in the long-run? Who knows, but what I do know is that learning keeps my mind sharp, opens new doors, and challenges me to think differently. And that’s a journey worth continuing.