Digital natives like Google, Uber, Amazon and Netflix have disrupted the IT landscape with state-of-the-art machine learning capabilities and are always on a look-out for advanced talent in artificial intelligence and ML. Now, the next-generation startups are launching products and services into the market and are addressing industry-specific pain points.
But as the ML ecosystem grows, the need for experienced experts has grown manifold. The buzz around ML has spawned a lot of online courses which serve as a good stepping stone for beginners, but the learning curve is mostly steep. Online courses from popular MOOC platforms do come in handy but enterprises and startups look for professionals who come with experience in ML technologies and have translated their work into a product or functionality. But as the tribe around ML is growing, so is the competition. So one of the key questions of our times is: How does one differentiate oneself from the crowd? Today, there is an oversupply of self-taught STEM learners, so one has to step up on the Kaggle and Github platform to differentiate.
It’s not enough to have theoretical knowledge to compete with your counterparts and get hired. Most organisations look for the best-of-breed, business-minded ML experts who can apply the technology to business problems:
According to Matthew J Schwartz of MJS Executive Search, a New York-based talent firm, top organisations lean towards ML candidates who have the following skills:
- A PhD in Machine Learning, Mathematics, CS in addition to business experience
- Knowledge and experience in leading edge and open source platforms
- ML experts should also have the ability to engage and consult with both business leaders and data scientists
- Researchers who have the most-cited papers, have their work published in journals, filed patents or clinched speaking engagements are preferred
- Have experience with cutting-edge ML technologies
- Deep Learning, AI-optimised hardware, decision management and biometrics
Those who are working at the intersection of engineering and big data ecosystem, are exposed to data modeling and at are building end-to-end production systems. Software engineers usually follow this route — software engineer < data engineer and usually deal with descriptive statistics. So, if you are a software engineer and trying to understand the ML ecosystem and plan to make a switch, here are a few pointers culled from community forums:
- ML candidate should boast of a strong coding background (especially Python) and also have a strong base in statistics, linear algebra and calculus
- Unlike big tech companies headquartered in the US, most Indian companies do not expect candidates to have published research papers
- What’s more important is the Github profile and even successful ML projects executed as a side gig for startups or small companies
- Another effective way of getting the word about yourself would be improving research around a real world application
- Also, domain knowledge is preferred for certain jobs, for example, a person from the NLP background would be expected to know the conversion of grapheme to phoneme conversion methods.
- On the other hand, if you are a complete beginner, try to look for mentorships and projects or repositories to start working on in your spare time
The Last Word
Professionals from data engineering and ML space are in high demand and companies are grappling to fill these positions. Finding and attracting ML and data science talent has become a strategic imperative for every company. Also, the hiring cycle is usually long and one of the reasons is a lack of clarity on the business problem, as observed by Facebook’s director of AI research, Yann LeCun. According to LeCun, who was cited in MJS Research, most tech companies are not able to find out the business problem and how to solve it.
We spoke to Senseforth CEO and co-founder Shridhar Marri who shared that ML beginners should go beyond their online learning and demonstrate the motivation with projects. “Online courses especially from international universities serve as a good starting point but it’s not just important to train, ML enthusiasts should also demonstrate their work with solutions. There is a lot of open source stuff out there and ML platforms that beginners can explore,” he said.