Job portal monster.com’s annual trend report has projected big data, AI and ML as the hottest job sectors in 2022.
Nitin Agarwal, Google Head of Cloud AI Industry Solution and Services (India), recounted the challenges he faced while hiring his team in India. “One common theme that I found, in the candidates that didn’t get selected, is that they prepared for the interviews well but lacked the “real” work… But the time I started having a detailed conversation on their projects, the problem starts coming up. Answers were very shallow and very textbook-ish,” said his LinkedIn post.
Though the demand for AI/ML roles is at an all time high, the niche talent is in short supply. A KPMG survey predicted 50% of the workforce will be preparing for AI, ML and related technologies in the next few years.
Of late, companies have started investing in their own employees to help them adapt to the latest technologies by putting them in upskilling and reskilling programmes. Experts believe incorporating AI/ML courses in the curriculum can make the workforce future-ready. However, with over 5000 engineering colleges still sticking with the traditional courses, the skill gap has increased in the industry.
Skillset difference
Data science is an umbrella term for multiple disciplines. While data scientists focus on algorithms and ensure the entire data processing pipeline is in order, ML engineers focus on the deployment of models.
A data scientist needs to have a deep understanding of a programming language, an IDE/visualization platform and a querying language.
Data Scientist is expected to be fluid in programming languages including Python and R. The goal is to ingest data, process it, feature engineer, build models and communicate results.
Data scientists also often use Jupyter Notebook or a popular IDE to code, write text, and view various outputs like results and visualizations from one place. Other popular IDEs include PyCharm and Atom.
Data scientists utilise structured query language (SQL) to query the first data, create new features, etc., after which the model is run and deployed.
Machine learning engineers come into play after the model has been built by the data scientist. They need to dive deeper into the code and deploy it.
Both data scientists and ML engineers are expected to know Python. However, machine learning engineers focus on more object-oriented programming (OOP) in Python, whereas data scientists are not as OOP-heavy. Most ML engineers also need to use Git and GitHub to version and store code repositories.
ML engineers are experts with deployment tools. There are plenty of tools like AWS, Azure, Google Cloud, Docker, MLFlow, Flask, and Airflow that ML experts are expected to know. Also, the title ‘machine learning engineer’ means ‘machine learning operations engineer’ (MLOps) as well in the job market.
While some companies prefer a well-rounded candidate capable of both data science and machine learning (operations), many prefer a specialist.
For candidates
The option of doing an added ML course from EdTech companies is always open. Companies always look for experienced candidates for ML deployments. Freshers find it hard to land big shot jobs in this area.
But candidates can overcome such limitations by demonstrating value via personal projects, open-source projects, hackathons, and coding challenges.
Wrapping up
The AI industry is rife with opportunities. However, the market is still nascent, with a high demand for a skilled workforce. Therefore, it is essential to put in the time, by both employers and employees, to bridge the skill gap and take the AI/ML industry to the next level.
