Making Machine Learning Accessible to Everyone: The BigML Story
Machine learning as a service (MLaaS) is a big deal in the cloud market. Forbes predicted that the global machine learning market would grow from $7.3 billion to $30.6 billion by 2024. To fuel this growth, data scientists and ML engineers are tasked with creating more models to meet the ever-dynamic business needs of customers and shareholders.
You don’t need to have a deep understanding of ML techniques to get the most out of them, which makes BigML stand out in the market.
Founded in 2011 by serial entrepreneur Francisco J. Martin, BigML is an ML service that offers an easy-to-use interface to import data and derive predictions from it. BigML lets you build machine learning and deep learning models without the need for coding.
In an exclusive conversation with Analytics India Magazine, Atakan Cetinsoy, VP, Predictive Applications at BigML, explained, “Our mission has been to make ML easy and beautiful for everyone. This contrasts with the complexities introduced by traditional ML tools, whether open source libraries or commercial tools. Our team’s foresight of the importance of the role ML can play for business has been beyond reproach, as the past decade has proven. »
Pioneer in the field
The first version of BigML was released in 2012. The team started with simple decision trees and added many more supervised and unsupervised learning models to the platform with each release.
Decision trees are a supervised learning method used to build a model that predicts the value of a target variable by learning simple decision rules from data characteristics.
“At the time, ML was primarily a subject of academic research and had yet to encounter the lexicon of business. As a result, most sales meetings in the early 2010s began by explaining what ML was. ML to Business Leaders BigML literally pioneered the ML-as-a-Service movement,” Cetinsoy said.
The platform’s easy-to-use user interface (BigML dashboard) enables code-free ML and requires little to no ML experience. “If you have experience, that certainly helps, but over the years we’ve seen many engineers and business experts embrace BigML, starting with free. educational videos and possibly continue with certificates to be able to build intelligent applications on top of the BigML platform. We also provide detailed documentation because we don’t believe in black-box ML approaches,” he added.
Built from scratch
BigML wrote all of its ML algorithms from scratch instead of assembling multiple open source packages to provide a smoother and more consistent end-user experience that is very robust. This approach has a higher initial product development cost.
“Yet after a decade and over a million lines of source code, we are in a very strong position with respect to the maturity of our platform and the overall end-user experience,” said Cetinsoy.
As a pioneer in the ML-as-a-Service space, BigML’s no/low code approach to ML eliminates barriers to entry for a much wider audience than the PhDs and data scientists that most other tools target. It provides a REST API that works well with programming languages such as Python, Ruby, and Java.
BigML offers many binding choices for developers to create custom workflows with their preferred programming language to solve a specific predictive use case. As an API-focused company, BigML has built one of the most comprehensive solutions ML-API where not only models but also datasets, assessments and many other platform artifacts are first-class citizens.
There are many tool vendors in the ML software space. In a sense, all ML platforms compete with non-consumers in the form of companies that have yet to deploy production ML systems. Next are in-house IT teams looking to build their own ML platforms, which is very expensive and risky unless you’re a tech giant.
Old guards like IBM, SPSS, and SAS have been catching up, looking to deliver ML capabilities in the cloud over the past few years. However, this can cannibalize their traditional statistical software business and has not been a radical transformation given the high costs and confusing product portfolios. That leaves VC-funded startups like DataRobot, RapidMiner, and Dataiku that rely on an amalgam of open-source libraries under the hood. As a result, their APIs lag behind. Some have struggled with the new macroeconomic backdrop given their high consumption rates.
“While I can’t provide a detailed comparison chart, I can summarize that BigML comes out on top because of its smooth learning curve, free level access that allows anyone to experience the platform, its flexible deployments and cost-effective private deployments,” Cetinsoy said.
Focus on the challenges of the current developer ecosystem, Cetinsoy said, “Developers have choices to make when it comes to building smart predictive applications. These range from already trained models exposed through APIs that can address a specific use case or data type (e.g. speech recognition) to open source tools and libraries with rather steep learning curves ( for example, Tensorflow).
The latest version of BigML is Object detection, which offers incredibly easy-to-use object detection capability. Before that, Image processing was released, allowing BigML users to treat images as another data type. The two recently released features are extremely useful for addressing a wide range of computer vision use cases such as medical image analysis, quality control in manufacturing, license plate recognition in transportation and people detection in security monitoring.