Data Science is progressing steadily in every sector, such as healthcare, pharma, retail, education, fashion, agriculture, etc. The increase in the number of IoT devices, which will be 20.4 billion by 2020, is mandating all of us to travel along the AI journey whether we like it or not. At the same time, it is presenting the possibility for collaboration.

Doctors are working with Data Scientists to improve patient care by automatically predicting diagnosis. Data Science algorithm for Fraud Detection is applied in various industries like fin-tech, real estate, and housing. Smart analytics and natural language processing assist customers in sectors like retail and fashion. With the convergence of commonality among different sectors and the adoption of AI steadily increasing,  the demarcation line among the roles and responsibilities of different professions are narrowing with more and more cross-collaboration. 

AutoML Democratize AI

Data Science applies machine learning to extract knowledge from data. Machine Learning involves supervised learning where models are trained with training datasets, unsupervised learning where we try to learn about the data structure without trained labels along with anomaly detection, artificial neural networks, reinforcement learning, etc.,

Applying Artificial Intelligence without the domain understanding leads to failure. At the same time, it requires time and effort for a person to adopt machine learning. AI revolution has given rise to AutoML, where machine learning models are readily available with the click of a button empowering people to become Citizen Data Scientist.

Anyone can become a Citizen Data Scientist

There has been steady progress in ML and AutoML since 2011. AutoML is encouraging Citizen Data Scientist from any part of the profession to apply Machine Learning to extract knowledge from the data that they deal with on a day-to-day basis.

The advancement in AutoML is enabling IoT devices with profound capabilities in security, education, entertainment, retail, fashion, agriculture, healthcare, etc., by inferring insights in the cloud or at the edge. 

The drag and drop feature allows non-technologists to conceptualize artificial intelligence and build products without coding. For example, Google’s Vision AutoML applies Machine Learning to classify any image collection in 3 simple steps.

Step 1: Upload Images

Step 2: Train images to be labeled 

Step 3: Predict label for the newly uploaded image 

AutoML reduces Time To Experiment

AutoML fastens the development of ML projects by automating feature engineering, model selection and hyperparameter optimization.

AutoML not only reduces the time to experiment different models and find the best-suited models, it also increases the accuracy of the models. Mercari, a popular shopping mobile app in Japan increased their image classification accuracy by 15% by employing AutoML.

AutoML fortifies IoT Development

The advancements in edge computing with edge analytics are converting real-time consumer data to intelligent insights. 

“By 2022, edge computing will be included in 40% of cloud deployments”

IoT along with AutoML is incorporating smartness into our lives by building smart homes, smart cities, smart grids, self-driving cars, etc.,

AI is still a Back-Box with AutoML

AI is still a black box miraculously solving problems for most customer segments. With AutoML, domain experts of any sector can adopt AI to improve their business without any prior knowledge of AI algorithms.

Even with AutoML, customers are still wary of the applications of AI solving the right problems. Domain Experts without the internals of ML algorithms fabricate incorrect models leading to more harm. There is a need to spread awareness about AutoML to not only employ AI across various sectors but to use AI for the right cause.

Hold-up Career Progression in Data Science

AutoML encourages anyone to become a data scientist. With the internals of data science hidden, AutoML is suited for beginners to quickly adopt AI.

At the same time, it does not provide the means to dig deeper into the internals of Artificial Intelligence to align one’s career and their passion.

Get Locked with specific AutoML Vendors

There are multiple vendors advancing in AutoML with H2O with their driverless AI, google with their cloud AutoML, DataRobot with their automated predictive models.

Not all vendors are open source, namely Google AutoML. With more and more new tools and algorithms arising in AI, it is not easy for all the vendors to keep up with the AI advancements. Most care should be adopted to employ the right AutoML for one’s use case and not get locked by using a specific vendor.

Bias & Awareness

AutoML automates the mundane process of cross-validating models and algorithms to find the best-suited models for a given use case. But sub-optimal models without the domain knowledge in fields like healthcare, fintech, pharma, there is a chance for bias unconsciously steering AI in the wrong direction. As of now, there is no easy way to incorporate bias monitoring to automatically detect if the training data is diverse enough.

With more citizen data scientists on the horizon, there is an immediate need for bias monitoring to spread inclusivity from the early stages of the AutoML revolution.


AutoML along with Citizen Data Scientists are proving that Artificial Intelligence is for all, paving ways to data democracy. At the same time, it poses questions making sure we are ethically conscious about using data for the common good.

As we are witnessing the next industrial revolution with AI, we are also responsible for making this world more inclusive by not violating the privacy and well-being of others.