How to Become an Analytics Engineer in 2026
Finding success and providing value in the age of AI
We can’t talk about analytics engineering in 2026 without mentioning the elephant in the room, AI. I’ve heard everything as dramatic as “AI will replace all data engineers and data analysts!” to “I don’t think AI will change a thing”.
I fall somewhere in between, because the truth is- AI will change a lot about how we work as analytics engineers. It requires us data people to adapt, but most of us are in this career because we are good at (and enjoy) adapting already. It’s hard to succeed otherwise.
If you’re thinking about becoming an analytics engineer in 2026, knowing how to properly use AI in your workflow can be a huge advantage, even over the most experienced engineers.
How can I bring value that AI can’t?
AI won’t replace you as a data analyst, data engineer, analytics engineer, etc. if you know how to use it to your advantage, and most importantly, if you focus on the things that AI is inherently bad at.
There are a few different things that I’ve found to be necessary for analytics engineers in the age of AI.
Focusing on foundational, core knowledge
AI is known to hallucinate, or give information that is straight-up wrong. If you don’t have a solid understanding of analytics engineering principles, you may not be able to discern the information that AI gives you. Building up a solid foundation of the best practices that have existed since the beginning of computers being invented will give you the base you need to succeed.
It’s time to pull out the good old fashioned books, blog articles, and newsletters for this!
Here are some good places to start:
Honing in on your soft skills
AI is a technical, logical system. While it is great at quickly finding and delivering information, it doesn’t have the capacity to put all of the information together in a strategic way. This is where you can shine as an analytics engineer.
Take the time to understand the business holistically and how your technical work directly impacts revenue. Practice explaining highly technical concepts to stakeholders in simple ways. When you can bring together technical knowledge with business acumen, you become unstoppable as an analytics engineer.
I recommend checking this out:
Producing QUALITY work over QUANTITY of work
You will never beat AI at the amount of work that it can produce. but you can beat it in quality. AI-generated code often lacks proper documentation and reasoning which eventually leads to tech debt. Instead, focus on the quality of your code through documentation, readability, and testing.
An article recently came out in the MIT Technology Review addressing this exact concern. The CEO of Sonar warns, “Issues that are easy to spot are disappearing, and what's left are much more complex issues that take a while to find. That's what worries us about this space at the moment. You're almost being lulled into a false sense of security."
Here are some resources for writing high-quality code:
All of this being said… you need to be an analytics engineer who uses AI in 2026.
Good news: AI won’t replace you if you take the time to build the skills I just mentioned. However, you do need to adapt and learn how to leverage AI in the work you already do. This means you need to take the time to understand prompting and how to work with AI to quickly get what you need.
Gone are the days of Stack Overflow. Instead, you have a super reader companion in your pocket that can give you the information you need in just a few types of the keyboard. Spend time asking it questions and using it as a helper in order to better understand what works and what doesn’t.
Lately I’ve been using Claude, Gemini, and Notion. Gemini and Notion are helpful for piecing together more general technical knowledge with internal company documentation or any special notes you’ve written for yourself.
Along with prompting, you should slowly add AI into your development stack. This means making the switch from VSCode to Cursor or nao, or installing the DataMates VSCode extension. These tools all bring AI into the same space where you write your code, helping you debug and perform monotonous tasks way faster than you could on your own.
By using AI tooling, you produce a higher quantity of work AND keep it high-quality.
Look out for a paid subscribers post coming soon on how to add AI-enabled tooling into your current analytics engineering workflow. Going forward, paid subscribers will receive a new technical deep dive post every month. I’ll focus a lot on AI in 2026 as well as foundational skills and new technologies.
I’m currently offering 20% off all annual paid subscriptions, so if you’re looking to up-level your learning in 2026 be sure to take advantage of this!
What foundational skills and tools do I need to learn to succeed in interviews and beyond?
If you had to start with 3 things I’d say SQL, data modeling, and a data transformation tool. These are the key skills of analytics engineering that you need to build before branching out into other sub-topics.
SQL
SQL is one of those skills that always pops up in an interview. When interviewing in the summer for a new analytics engineering role, I found companies leaning away from the LeetCode/HackerRank problems. Instead, they asked me questions focusing on their business using real company data.
The best way to prepare for this is through actual business case studies. Data Pipeline Summer walks you through how to create a data pipeline using data from a fake company called TrailTrekker. Complete the weekly challenges and you’ll gain the extra SQL practice that you need to succeed.
If you’re a complete beginner, start here with SQL basics.
Data Modeling
It’s usually a good sign if a hiring manager tests you on data modeling principles. This means they understand the foundational work that needs to be done to power an analytics team. Whether they ask you these questions in a spoken interview or include them as part of a take home project is dependent on the company.
To properly prepare, familiarize yourself with Ralph Kimball’s fact and dimension models as well as snowflake and star schemas. I highly recommend reading the 7 Data Modeling Concepts You Must Know to better assess where to begin your studies.
dbt and/or SQLMesh
While dbt is more popular, and these tools have recently merged under Fivetran’s umbrella, the principles behind both are the same. They help you write modular, well-tested, well-documented SQL code. While not all analytics engineering roles require this, most of them do.
Most companies, if not currently using a data transformation tool, will want to in the future. Understanding the best practices behind them will give you an advantage. I teach you about both dbt and SQLMesh, including project setup, in Data Pipeline Summer.
Airflow (or another orchestration tool)
Orchestration and data modeling are closely connected. While orchestration more so focuses on how data models interact, it’s important that you understand it so you can write clean DAGs. While many data teams use dbt Cloud for a simple form of orchestration, many are using tools like Airflow or Prefect.
Marc Lamberti is hosting a free webinar later in January covering the Airflow 3 certification exam. Whether you want to get certified or not, this is an excellent free learning opportunity to learn more about tasks, schedulers, and DAGs.
I touch on all of these skills, tools, concepts, and more in my ebook The ABCs of Analytics Engineering. This is a great place for anyone to start if they are new to this career path or the more “modern” world of data.
How can I use AI to improve my resume?
You’ve most likely seen videos of people who use AI tools to scan job descriptions and generate their resumes based on these. I’ve heard of people doing this, using these bots to apply to thousands of open roles a day.
I DO NOT believe in this. One, I don’t think it actually works, as these are the people that often remain jobless for months. Two, lying on your resume will cause you more pain than its worth. After all, you still need to nail the interviews and work the actual job…
This being said, I recommend having a solid resume of your prior experience with actual tools, skills, and projects listed. Then, when you come across a job description that aligns with your experience, you can use AI to tweak it and feature some of the same keywords.
For example, if I have a bullet point on my resume that says, “Increased data availability by 30% by using Prefect to orchestrate my dbt data models” and a job description is looking for someone who has “experience building data pipelines using Airflow or similar tool”, I may change my bullet point to say, “Increased data availability by 30% by building a data pipeline using an orchestration tool similar to Airflow called Prefect”.
“Increased data availability by 30% by using Prefect to orchestrate my dbt data models” —> “ Increased data availability by 30% by building a data pipeline using an orchestration tool similar to Airflow called Prefect”
I’m not lying about my experience, I’m just altering my resume to include the keywords that a recruiter may be looking for.
How can I use AI to excel in non-technical interviews?
Using AI to help me prepare for potential interview questions was hands-down the most useful way I used AI in my job search. I gave Claude the job description and the company’s documentation on the format of the overall interview process and asked it to come up with potential interview questions.
The questions were all over the place from technical questions to work style questions, but gave me a great diversity in what I needed to prepare for. Most of the times, they were spot on! Whenever a question in an interview surprised me, I added it to a document and used that to inform future prompts.
Once I had these questions, I wrote out some of my major accomplishments/failures using the STAR method. This allowed me to then have a few different situations I could reference depending on the question. Remember- you don’t have to come up with a million different project summaries, but really a few that touch on different key points.
Another thing I’ll add about interviewing- make sure you also interview the people who are interviewing you. You want to work with people you like! If you don’t like someone during this process, you most likely won’t like working with them
I’ll be sharing an in-depth guide on using AI to job hunt including prompts and my master interview question sheet later this year.
If you’re looking to become an analytics engineer, or become a better analytics engineer in 2026, you’re in the right place. I’m so happy you are a part of the Learn Analytics Engineering community!
I can’t wait to see what this year has in store for us all! 🫶
Madison
Brilliant breakdown of where analytics engineering is heading. The shift from LeetCode-style problems to business case studies actually mirrors what I've seen in production codebases too, where context matters way more than algorthmic tricks. The point about AI lulling teams into a false sense of security is spot on, I've debugged so much undocumented AI-generated code that looked perfect at first glance but had subtle logic errors baked in. Quality documentation becomes the real differentiator wen everyone can generate syntactically correct code.