Table of Contents
What is Machine Learning Engineer?
What does a Machine Learning Engineer do?
Machine learning engineers (MLEs) work on the platforms and infrastructure that facilitate the development, deployment, and monitoring of machine learning models. They might also assist with implementing and scaling Machine Learning models for production or work directly on application-specific Machine Learning systems in a role comparable to an Applied Scientist. Some examples of the platforms that help with the deployment of machine learning models are:
- Michelangelo by Uber
- Metaflow by Netflix
- SageMaker by Amazon
- IBM Watson Studio by IBM
- Azure Machine Learning Studio by Microsoft
Most machine learning engineers have previously worked as software engineers and have strong expertise in DevOps and software development. Machine Learning Engineers often use more enterprise-oriented and performant programming languages to create systems rather than scientists who utilize Python.
There is a stronger focus on technical design and architecture, infrastructure, scalability, and security. Machine Learning Engineers, who are essentially software engineers with an emphasis on Machine Learning, are the ones that put up the infrastructure and processes around it in contrast to scientists who mostly use technologies.
The table below highlights the differences between Research Engineers, Applied Scientists, and Machine Learning Engineers:
Source: Eugene Yan
Machine Learning Engineer career paths
Since Machine Learning Engineers have a wide variety of expertise, they can branch out into different roles depending on their personal goals and interests. Machine Learning Engineers can pivot into new roles if their current role isn’t as exciting as expected. Some additional career paths for Machine Learning Engineers include:
- Data Engineer: Data engineers create and maintain the data platforms on which machine learning and AI systems rely. Their primary responsibility is designing information systems for activities including data collection, processing, conversion, mining, and pattern recognition.
- NLP Scientist: NLP scientists aim to create and build tools and software to recognize human speech patterns and translate spoken words into other languages. The idea is to make it possible for machines to comprehend human languages as people do. A strong command of at least one human language and an understanding of how machines work is prerequisites for this machine-learning vocation.
- Machine Learning Cloud Architect: The key responsibility of a Machine Learning Cloud Architect involves overseeing an organization's cloud platform. Experience in architecting solutions in AWS and Azure and knowledge of configuration management systems like Chef/Puppet/Ansible are among some of the required skills for cloud architects.
Leading companies to work for as a Machine Learning Engineer.
Some factors to consider when evaluating potential employers include the company's mission and values, the support and resources available to scientists, and the company's culture/work environment. Companies at the cusp of technological innovation and working on cutting-edge projects will have some of the best resources and opportunities for Machine Learning Engineers.
Here are a few examples of companies that are often considered to be among the best places to work for Machine Learning Engineers:
Machine Learning Engineer Salary Components
Depending on your location, the base salary offered by companies like Apple, Facebook, and Amazon will differ, along with the rest of the compensation components. Most companies compensate Machine Learning Engineers according to the cost of living in their respective location and the going market rate for engineers in the area, so it's essential to be mindful of how your location will affect your pay. If you are not in the Bay Area, NYC or Seattle, your comp will likely be lower than the numbers you find online.
With that being out of the way, the base salary at Amazon and Apple, unsurprisingly, is the component that moves the least when negotiating. The aforementioned companies have a small band for the base salary within each level, so you'll likely only see your base move by a maximum of $20k-$30k.
Annual bonuses are another component that can be coupled with the base salary. The annual bonus is usually non-negotiable and highly subjective to the company you are applying to. Most big-tech companies offer performance bonuses; it is essential to be aware of the annual bonus when negotiating, especially when you have multiple offers. Companies such as Google and Facebook often apply a company multiplier to performance bonuses, which they also do for stock refreshers.
Of course, annual bonuses are non-negotiable, but if negotiating with a company like Amazon that doesn't offer them, you can always factor them into your counteroffer's base salary as your "yearly cash amount.”
Companies like Facebook and Microsoft follow a standard and linear vesting schedule of 25% yearly (typical initial grants last four years). Although, companies are starting to get very creative (to gain the upper hand) in how they vest RSUs.
At Facebook, Microsoft, Apple, and most other companies, RSUs are subject to a 4-year vesting schedule: 25% vests at the end of the 1st year (sometimes accompanied by a cliff), then 25% in each of the 2nd, 3rd, and 4th years. For example, if you were given a stock grant of 800k at Apple over four years, the equity would vest as follows:
- Year 1 - 25% $200k
- Year 2 - 25% $200k
- Year 3 - 25% $200k
- Year 4 - 25% $200k
Google and Amazon do equity vesting a little differently. Google has a front-loaded equity vest (33-33-22-12), while Amazon has a back-loaded equity vest (5-15-40-40). They often use these vesting schedules to inflate your offer and make it look more substantial than it is. In Google’s case, the recruiters often quote the first-year number as your per year total compensation, while Amazon adds a ‘conservative’ 15% growth factor on your future equity. Knowing how the recruiters frame your offer is paramount, as it could be the difference between accepting a great offer and an outstanding one.
We have seen Machine Learning Engineers get offered a small signing bonus without asking for it - recruiters often use this tactic to sweeten the deal. It’s often possible to increase the signing bonus substantially (this is where having leverage helps!). With the proper leverage and framework, Machine Learning Engineers can get up to $100,000 in Tier 1 locations (Bay Area, New York), which holds across companies like Apple, Google, and more!
Albeit, it's a common recruiter trick to leave a signing bonus out of the initial offer so they can add it as part of the negotiation (and avoid increasing the base or equity, which could require more senior-level approval). Some recruiters have claimed that the company does not offer sign-on bonuses, which isn’t always the case. Recruiters at Apple and Facebook will not initially include a sign-on bonus. Instead, a sign-on bonus is often added when specific leverage/information is shared with them.
Most companies will reserve the right to “claw back” a portion of your signing bonus if you leave before the 1-year mark. This is normal for major tech companies, commonly only requiring you to repay the pro-rata amount — for example, if you leave after ten months, you would need to pay 2/12 of your signing bonus back.
Recent Top of Band Machine Learning Engineer Offers
These are some of the offers we helped negotiate in Tier 1 locations like SF and New York. Although the numbers mentioned above seem very enticing, the proper framework and leverage were used to achieve such strong results. According to our data, Machine Learning Engineers earn about 15-20% more than software engineers on average.
Machine Learning Engineer Salary Negotiation Process
Before preparing for a negotiation, make sure you have a good understanding of both your financial and career goals. This will help you decide what you should be asking for and make you better prepared to negotiate effectively.
For example, if you hope to save a certain amount of money each month or save for a specific goal, you should make sure the salary you are negotiating will help you get there. This may mean you ask for more salary instead of a non-cash benefit like equity.
An explicit goal of what a pay increase will help you achieve will help you feel more confident to ask.
If you haven’t yet received an offer, here are a few things to consider during the interview process:
- Do not share your current compensation. In many states (e.g., California), it is illegal for companies to ask for this. If a recruiter asks you, you are certainly within your rights to say, "I don’t feel comfortable sharing that information."
- Related - we do not recommend sharing your compensation expectations before receiving an offer. Most companies will pay very competitively and will be willing to negotiate after giving an initial offer. If you choose to throw out a high number when asked, that will increase the chance you are required to provide proof of a competing offer. Instead, if you’re asked for your pay expectations, we recommend you reply with, "I'm focused on the interview process and still researching market data. I am confident we will get to a number that works for both of us."
Negotiating a salary, equity, and signing bonus for a Machine Learning Engineer offer can be daunting. Still, with the proper knowledge and preparation, you can increase your chances of securing a fair and competitive offer. We recommend you:
Machine Learning Engineer Negotiation Tactics
Recruiters commonly use a handful of sneaky tactics to help pull the negotiation in their favor — and it’s essential to be aware of them to avoid being taken advantage of. The most common tactics include putting time pressure on you with an exploding deadline, mentioning that the initial offer is non-negotiable (even though it is!), selling you on company growth and saying that your equity value will increase substantially, and promising to revisit pay shortly.
Some of the most common negotiation strategies that we use in rebuttal are:
- Putting pressure on the employer – Employers are often pressured to fill a position quickly. You can use this to your advantage by pressing them to offer you a higher salary by saying, “I know you’re trying to wrap up the negotiation - here’s what I’d need to sign.”
- Standing in a stronger position – If you have a competing job offer (or are currently employed), you are in a stronger negotiating position. You can use this to your advantage by asking what you want regarding salary, benefits, and equity.
- Sharing outside information – You can also use outside information to strengthen your position. For example, you could talk with a competitor about the compensation they might be offering. Recruiters can be great resources for this information, too!