The Most Practical Way Into AI? It Might Be Machine Learning Engineering

Table of Contents

---

• First, Let's Untangle the Job Titles
• What ML Engineers Actually Work On
• A business-first mindset will set you apart
• The 3 Mistakes Early-Career ML Engineers Make
• What the Field Looks Like Right Now
• How to Get Started

Machine learning engineering is part software engineering, part data science, and entirely focused on getting AI models out of research notebooks and into production systems that real businesses actually run on. The role doesn't get as much attention as "data scientist" or "AI engineer," but it may be the most practical entry point into the field right now – especially as companies shift from experimenting with AI to actually deploying it. We spoke with Artem Yankov – a former engineer at Zillow, Rover, and Google, and now a mentor for Springboard's Machine Learning Engineering and AI Bootcamp – to find out what the role really looks like from the inside.

First, Let's Untangle the Job Titles

Before diving into what ML engineers do, it helps to understand where the role sits relative to adjacent titles you'll see on job boards.

A data scientist typically focuses on analysis, experimentation, and building models – asking questions of data and generating insights or predictions. A data engineer is more focused on the infrastructure that makes data usable: building and maintaining the pipelines that move, clean, and store data so others can work with it. A machine learning engineer sits closer to the software engineering side – their job is to take models that have been developed (often by data scientists) and integrate them into production systems reliably, scalably, and efficiently.

As Artem Yankov – a former engineer at Zillow, Rover, PNNL, and Google, and now a mentor for Springboard’s Online Machine Learning Engineering and AI Bootcamp – explains it, the shift in his own career illustrates this well: "As tools like AutoML started to abstract away a lot of model building, my work naturally shifted toward integrating models into data pipelines – that's how I ended up as a machine learning engineer."

In other words, ML engineers are less focused on developing the algorithm from scratch and more focused on making sure it actually works in the real world – at scale, with real users, under real constraints.

What ML Engineers Actually Work On

At Rover.com, Artem built ranking algorithms: the systems that decide which pet sitters you see first after entering your preferences. At Zillow, his work involved pricing the ad slots where real estate agents' faces appear next to listings. At Google, he worked as a data engineer on the satellite imagery team, helping turn raw satellite images into something usable in Google Maps and Google Earth.

These examples illustrate something important: ML engineering often doesn't look like the dramatic AI you read about in headlines. It's ranking algorithms, pricing models, recommendation systems, and image processing pipelines. It's the unglamorous work of making predictions useful and reliable at scale.

A business-first mindset will set you apart

Technical fluency matters, but if you can't connect your model to a business outcome, you haven't finished the job. Developing this instinct early is one of the most valuable things you can do as an aspiring ML engineer.

In practice, that means understanding how companies evaluate ML investments. Most start with an A/B test: one group sees the new machine-learning–powered feature, another sees the original, and the team tracks business metrics – customer acquisition cost, revenue per session, customer lifetime value, churn reduction. "Even when a model 'works,' it must outperform the operational cost of running it before it truly delivers ROI," Artem says. That means accounting for retraining cycles, monitoring for model drift, cloud compute costs, and the engineering hours required to keep it stable.

Latency is another business consideration that catches newcomers off guard. A model that technically improves outcomes but adds too much lag to a checkout flow can hurt revenue rather than help it. "Models can slow down processes, negatively affecting customer experience, especially in fast-paced environments like online shopping, where speed is crucial for competitive pricing," Artem notes.

The takeaway: every technical decision has a business consequence. The ML engineers who thrive are the ones who internalize that early – not after years on the job, but from the very first model they build.

The 3 Mistakes Early-Career ML Engineers Make

Ask Artem what separates the practitioners who thrive from those who struggle, and he comes back to three patterns he sees repeatedly.

1. Forgetting about business impact. "Many practitioners, especially those with PhDs, become overly focused on technical details and algorithms, forgetting that machine learning must serve the business," he says. The model is a means to an end. If you can't articulate how it moves a metric the business cares about, you haven't finished the job.
2. Chasing technical perfection. "It's often better to launch a functional product quickly than to spend months striving for a slightly better model," Artem says. His advice: always start with the simplest possible baseline – even a rule-based system – before building something complex. A sophisticated deep learning model might offer only marginal improvement over that baseline, making the added complexity and compute cost hard to justify. "Tools like AutoML, especially from cloud providers like AWS or Azure, make establishing a baseline easier."
3. Picking the wrong metrics. This one is subtler but consequential. "Using 'accuracy' to predict a rare event like cancer is misleading," Artem explains. "A terrible model that always predicts 'no cancer' will appear highly accurate." Choosing the right evaluation metric has to happen before any model is built – and it has to be tied to what the business actually needs to measure.

What the Field Looks Like Right Now

Artem doesn't sugarcoat where he thinks the AI market is headed. "Many companies are struggling to extract real value from AI developments," he says, drawing a parallel to the "AI winters" that followed previous cycles of hype. "Given the amount of money poured into the space without seeing ROI, and the current energy infrastructure limitations, I predict some air will be taken out of the bubble."

But he sees that as an opportunity rather than a threat. "I think the most useful near-term role of AI will be helping businesses implement traditional machine learning techniques more effectively. You don't have to deploy a giant AI model to see value. You can use AI as a helper – to design experiments, pick metrics, debug code, or understand how to integrate simpler models into your data pipelines."

That's precisely why he believes ML engineering skills remain durable even in a cooling market. "The need for engineers who can evaluate models, measure impact, and build clean data pipelines isn't going anywhere."

How to Get Started

For anyone considering a path into ML engineering, Artem's most practical advice is to start on Kaggle.

"Kaggle hosts real datasets and real problems from companies, and you can experiment with building models, see how they perform against others, and learn extremely quickly. The forums are also incredibly supportive – people share starter code, troubleshooting tips, and walk through their approaches."

One caveat: "The top Kaggle solutions are often too complex to ever be deployed in a real company. So treat Kaggle as a learning sandbox, not a template for production systems."

On the math front, he's reassuring for anyone who's been intimidated by the technical prerequisites. "There is a lot of math behind machine learning, but most of it is abstracted away – you don't actually see it day to day. For most students, the math you need is pretty basic: addition, subtraction, understanding ratios, reading metrics, interpreting plots, and some light statistics like correlation or p-values. You don't need linear algebra or advanced calculus to succeed."

The most important thing, in his view, is getting started and staying focused on the practical. "Don't be intimidated, and don't get distracted by the hype. Whether AI continues accelerating or cools off a bit, companies still need people who understand how to integrate models responsibly into real systems."

Find out more and read Springboard reviews on Course Report. This interview was produced by the Course Report team in partnership with Springboard.