If you think that your company could benefit from Artificial Intelligence and Machine Learning applications, the first step is to find a team that can help you solve your problem. The following are six questions you should ask a Machine Learning consultant to determine if they’re the right team for you:
1. Do I even need Machine Learning?
The truth is that AI, Machine Learning, or any other buzzword you may use, is not the cure for all of the world’s data problems. Run, don’t walk, away from anyone who promises that it is.
When implemented properly, Machine Learning plugins can increase the productivity of your employees by 2x or 3x. These results have been achieved by many of our clients and in this article, we will show examples of how it can be done.
We like to think of an AI system as a friendly sidekick that can solve a clear task, such as automatic article recommendation, detecting objects on an image or financial risk prediction.
A good Machine Learning consultant will first assess whether you actually need Machine Learning. They’ll determine whether you’re likely to get a return on your investment (ROI).
We speak to a lot of people and quite often we talk them out of building an AI solution – even though it might mean business for our company. For instance, early-stage startups don’t need AI until they reach a critical mass of users.
A good rule of thumb for whether you need an AI solution is: do you have a well-defined data processing problem that is important for you to solve? Can you quantify the benefits (monetary or otherwise) of solving it?
If you’ve established that building an AI solution is indeed for you, read on.
2. Is your expertise a match for my problem?
There are various domains where Machine Learning can be applied. Working with images and text requires different expertise and a different approach. While some algorithms/techniques overlap and are domain agnostic (e.g. LSTM design can be used to work with both text and image sequence data), there are substantial differences depending on whether you’re working with tabular data, text or images.
For example, a good Machine Learning specialist for image processing should also understand the general challenges with computer vision – he/she should know when to use OpenCV versus Deep Learning, how to observe underfitting/overfitting on the evaluation dataset, and how to perform augmentation.
However, a robust Natural Language Processing (NLP) specialist needs to understand things like how grammar works, how text is composed, and what embeddings are and how to build them. He also needs to be familiar with stemming, recurrent deep learning algorithms, and boosting techniques.
3. Do you have general programming skills? Can you build, deploy and maintain a scalable solution on the Cloud?
It depends on your needs, but it’s somewhat likely that after you build your Machine Learning model, you’ll want to be able to plug it into your existing application. Deep Learning requires GPU to run efficiently, and scaling models requires expertise.
A lot of Machine Learning consultants have a solid math background, but not much expertise in large-scale, or GPU programming. It’s better to be a specialist than a generalist, but you’ll need someone to build end-to-end Machine Learning pipelines. You can read our Case Study to understand how we build such pipelines as well as see some code samples.
4. Are you up-to-date on the latest, cutting edge scientific literature? Do you have expertise in Deep Learning?
While Deep Learning isn’t always applicable (and arguably not even in most cases), it is an essential technique to consider especially when working with unstructured data.
We define “unstructured data” as “easily understandable by humans, but not by computers” – examples would be images or text. While Deep Learning has delivered groundbreaking results and almost entirely solved tasks like image classification or object detection, it still requires serious expertise to work with.
5. Is your experience commercial or academic?
Machine Learning is by far the most heavily researched field of Computer Science at the moment. It has brought together the world’s sharpest minds – the brightest mathematic and computer science researchers – to focus on this one area for the foreseeable future.
While strong technical skills and analytical thinking are essential prerequisites, there’s more required than that. Not many ML scientists have had an opportunity to work on analyzing, solving and successfully deploying a Machine Learning-based solution for a client.
An ideal ML consultant has experience in approaching problems with both scientific curiosity and business dexterity.
6. Does your consultant take security & encryption seriously?
Is your firm processing sensitive data worth hundreds of thousands of dollars? Does your project involve working with government data? What would it mean for your firm if that data was compromised?
A system is only as strong as its weakest link. This means you need to do everything you can to protect it.
For example, if a consultant considers SSL as the only requirement necessary to achieve security, you should dismiss them immediately. SSL is just the starting point (along with VPN, database encryption, monitoring & auditing) – you need much more advanced methods to achieve reliable security and every good consultant knows this.
You also need strict, relentless people procedures to increase your chances of being safe.
When it comes to data, there’s no simple one-size-fits-all.
It’s hard to find one person that can have all qualities, and that’s why we assembled a team to solve the world’s most challenging data problems, and to a great success.
If you’re curious to learn more about Sigmoidal, click here.