A partnership between data specialists Knowledge Integration Dynamics (KID) and global AI Cloud leader DataRobot is automating and democratising artificial intelligence (AI) and machine learning (ML), putting it into the hands of more South African businesses.
Markus Top, who is heading up the partnership at KID, says it is a logical next step for KID, which has supported South African enterprises through their data journey for over 20 years. “Every business today wants to be data driven and embed AI at scale. However, until fairly recently achieving this has been a costly and time-consuming task. With DataRobot, the manual, time consuming processes within AI and ML projects are largely automated, allowing businesses to transform and innovate faster.”
DataRobot democratises AI through machine learning automation (AutoML) and extensive machine learning operations (MLOps) capabilities, making it easy to embed ML into business processes at scale, with best practice and guardrails in place. For business, this means faster time to insight and a competitive edge, for IT and DevOps it means it’s easier to get ML models into production and monitor them, and for data scientists and subject matter experts, it means speed, impact and scalability of solutions.
DataRobot also delivers centralised monitoring, management, governance, and alerting with a Continuous AI framework with automated retraining, and challenger models for model comparison. DataRobot also allows data scientists to build models using their preferred tools and languages and deploy them to the DataRobot platform.
DJ Human, senior data scientist at DataRobot EMEA, says: “Many tools claim to do what we do, but DataRobot is a pioneering voice in an AI, having been in the industry for 10 years with the most complete solution for the automation of machine learning development and deployment, with governance and guardrailing end to end.”
Johann van der Walt, Director of Operations at KID, says: “This partnership brings to South Africa a very niche, focused set of AI data solutions. DataRobot is one of few platforms dedicated entirely to AI and ML. From a maturity point of view, it is at the front of the pack. And KID focuses purely on the data management lifecycle: over 23 years, we have naturally grown into what the ‘big next’ is for industry. When it was integration, MDM, data quality, governance, predictive analytics, we were there. Now, together with DataRobot, we see the need to democratise data solutions to meet changing business needs, and we have moved to meet this need.”
Van der Walt says South African organisations are seeking improved predictive capabilities, often to support compliance and risk management. “While banks and financial institutions were early adopters, we now see organisations across sectors looking to use AI and ML to achieve efficiencies, market share and a competitive edge, and companies want to move forward faster than competitors,” he says.
Putting data scientists higher up the value chain
Human says DataRobot not only speeds up project development, it also allows organisations to optimise their scarce data science resources.
“Demand for data scientists far outstrips supply; and when organisations have smart and capable data scientists who could do a lot for the business, they tend to get sucked into coding all day. They often get stuck in months of development cycles. However, when repetitive manual tasks are automated, data scientists are freed to use their expertise for problem framing to drive business value,” he says. DataRobot offers a web-based visual user interface, as well as API access that code-first users can leverage through their preferred IDE or notebook tools.
Janco van Niekerk, data scientist at KID, says AI and ML projects can be tedious work – particularly for highly skilled and qualified data scientists. “These projects can be very time consuming: you may first have to make sure everyone understands what AI and ML are, the business comes with ideas, then you meet in the middle and decide on an approach. Then after that, there is a lot of coding, with months of manual, tedious and repetitive tasks to deliver a ML project. But with DataRobot this all becomes shockingly easy,” he says.
Van der Walt notes that the massive data scientist shortage has resulted in headhunting and movement; which drives costs up. “DataRobot enables organisations to deliver a great AI or ML service without having to find more data scientists with master’s degrees and doctorates.” Human adds that: “As more young data scientists enter the field, DataRobot will empower them to take charge of projects and free up senior staff to address the more complex projects. This will change the profile of the skills required and over time, it will help address the skills shortages.”