Machine Learning and Data

The prerequisite to using ML/AI is having data. Data is used to identify trends and problems and act on them to improve the user experience of an application. Machine Learning takes all the data available and does everything that a human being would do in their mind in the process of figuring out a solution to an issue, except it does it faster and better and at a much larger volume. It produces results by doing all the behind the scenes work that you would have done manually.

The biggest issue with data is collecting the data, as there are many questions that arise in the process of doing this. There is data flowing through every organization but what is the right amount of data? What should the data look like?  Often you’ll notice that there are missing values, how do you input that? Then you want to make sure the data is representing the machine learning problem correctly. You don’t want to generalize it too much otherwise, it won’t solve the problem.

Often too, there are questions about the security and privacy of persons as companies mine all this data. It might be helpful to know that usually, companies use metadata, and not the individual’s personal data to acquire the answers that they need to improve their product. However, security and privacy concerns are common and valid concerns in the field of Machine Learning.

Evidently, questions are not unusual in the ML/AI world. In addition to those mentioned above, you will have questions when it comes to model performance too. A model is a file in ML/AI that has been trained to recognize certain types of patterns and make predictions. To use it, you’ll need to ask yourself: How does a model need to perform for you to know that it’s the best model? Or how often should you retrain your model? Then when deploying the model, you’re faced with questions regarding how to take it to production or how often you should retrain the model.

So if you’re looking to work in this space, don’t be afraid of questions. Don't be afraid to ask them, to receive them, and to work on finding answers to them.

Breaking into the Machine Learning Space

Aniththa’s academic background is in Computer Science and Finance and so to better understand what was expected of her as a Product Manager working in ML, she took some relevant courses to acquaint herself with the basics. She advises aspiring Product Managers going into a sector that they’re not familiar with, to engage in a knowledge-share with other Product Managers or even Engineers on their team, and understand how they learned.

You’re working in this new, niche sector but remember that you’re first of all, a Product Manager. That means you’re responsible for bridging the Machine Learning gap as well as the gap between customer expectations and engineering. You need to understand the product, product goals and customer. In fact, Aniththa recommends that the Product Manager be customer obsessed. Understand the user experience and what the customer needs so that you can communicate these needs to the data scientist, then get information back from them and communicate that to Engineers and customers.

Understanding the entire Machine Learning process is integral to doing your job. You don’t need to become an expert in the field but understanding the flow helps you know whether there are any missing pieces.

If this sector is of interest to you…

Find opportunities to build relevant experience especially when you’re building your resume. Tailor your skillset to being a Product Manager by using programs like Co.Lab that give you relevant experience. Even if your product isn’t successful at the end of the day, you’d have learnt something valuable.

Seek out mentors to guide, empower and teach you especially if you’re new to the space and stay in touch with your mentors for guidance and access to opportunities. The keyword here is to seek out mentorship, don’t wait for it to come with you.

Understand that your primary assignment is to advocate for the customer. As Machine Learning is a new field, try understanding the process for yourself first by working with stakeholders and understanding their processes.

Finally, as someone with experience working in three of the biggest tech companies in the world, Aniththa wants you to remember that a company has a vision, and you must ensure that your vision/career goal lines up with that of the company and lines up with the team you’re in and that way, you’ll find value in your work.

To watch the entire discussion with Aniththa, from which this article was derived, visit our YouTube channel. Also, stay tuned to our socials to join in on our next discussion and visit our website to find out more about Co.Lab!