Why “learning” to hire for Machine Learning experts is what your startup needs

Machine Learning experts

Companies like Google and startups like Clarifi have one thing in common: they have identified the highly evolving symbiotic relationship between their need and importance of exploiting the power of systematic machine learning approach. For example, Google’s Translate and Photos along with Voice Search is a result of applying machine learning on existing domains. Clarifi has used machine learning to develop software and understand videos, which can add value to their company as an advertisement tool. This relationship has naturally altered the hiring process for ML experts. Companies today face immense competition (especially in the Silicon Valley) to find and attract new talent in the Machine Learning and Data Sciences domains to be ahead of competition. This in turn has impacted factors that Hiring Managers need to consider while scouting for, interviewing and having on board their most-poached category of employees, the Data Scientists. Here are the big three factors that recruiters and hiring managers would have to deeply ruminate about when it comes to Machine Learning openings.

1. Narrow down what ancillary skills your Machine-Learning expert must-know

While hiring Machine-learning experts you must ensure that they would be able to work closely with business stakeholders in order to understand your specific business issues and accordingly develop the data driven insights. It is therefore apt on your part to prepare a small dossier to make the person understand your exact requirements. For that you must yourself learn from your internal teams the requirement so that you are able to assess the candidates. For example, is he/she expected to develop data ingestion models from different clients’ touch-point sources with appropriate text mining analysis? Is there any need to audit the manual or automated tagging process to ensure the viability of the data that would be presented in the systems? Is he/she adept in analyzing demography-specific data and deliver authentic insights to the team? Is the scientist capable of exploiting tools like MS Access, SQL, Excel and presentation tools like Tableau (or the one your organization uses)? Will he/she be able to develop new metrics and recommend strategies to improve business results? Other than these technical questions you may also ask them on their favorite data scientists or the startups to see if that matches with the vision of your company.

2. Determine for how long you would need a Machine-Learning expert

The role of a hiring manager in a startup will be different than if he were to hire for an established company. He should aim at bringing in a machine-learning expert who will ensure optimum usage of existing resources. As a hiring manager, you need to have a clear idea about the kind of machine learning projects underway and its requirements. If it is a new project then a completely new set of machine learning algorithms needs to be designed by the expert, which requires you to hire an expert for a longer duration of time. You also need to determine the extent to which the expert will be allowed to perform. There are two ways to do this – they can build the core algorithm and deliver it or they can convert that into deployment language, besides developing and deploying the algorithm. If conversion and deployment by developers seem to be more cost effective, then you can let go of the machine-learning expert as soon as they have delivered the core algorithm. To save further costs you can let data managers to prepare the data to be used by your machine-learning expert. All these decisions will depend on what kind of project your company is getting involved in.

3. Customize your psychometric assessments for hiring Machine-Learning experts

Psychometric assessment during recruitment only works if the employer knows what they want. Now, when you have identified the skill set and nature of employment for the machine learning expert you need to develop an ideal questionnaire that would discover if the person is ‘at tune’ with your company’s vision or what motivates them. You can also identify inconsistencies and judge how the person is going to behave under pressure or real life situations. Since psychometric and algorithmic assessments apply statistical models backed by machine learning to analyze the candidate information, the process is more or less foolproof. Then there are platforms like Kaggle for predictive modeling and analytics competitions that recruiters are using to dig in on a great volume of data set. Ironically, Machine learning, which is the newest weapon in the arsenal of data science, is now being used in recruiting the experts in the field!

Since Machine learning is to “…focus on deep learning, natural language processing, image applications, recommenders, physical sensors and signal processing…”[Source: Apple] it is not as simple as hiring a programmer or an ERP expert. Recruiters should look out of the box and think of machine-learning experts who can work on open source applications like R. This will not only save costly investments in computing platforms but will also make the process flexible.

ShareShare on FacebookShare on LinkedInTweet about this on TwitterEmail this to someone

Leave a Reply

Your email address will not be published. Required fields are marked *