Lana platform recently conducted an interview with Kurt Muehmel, Head of AI Strategy, and Sid Bhatia, Area VP & General Manager – Middle East, Turkey & Africa at Dataiku. They discussed Dataiku‘s philosophy, its rapid growth, and the intricacies of integrating practical enterprise AI with governance and technical controls. They also detailed the challenges and successes of real-world projects with major clients, particularly in the Arab region and globally.
When asked about the nature of Dataiku, Kurt Muehmel stated: “It is a comprehensive artificial intelligence platform, described as a ‘global system for artificial intelligence’ used by major institutions worldwide, including banks, industrial companies, insurance, retail, and even healthcare organisations. The platform enables the connection and integration of human expertise (business experts and operations teams) with data analysis teams and artificial intelligence experts, within a flexible interface that facilitates analysis, solution building, and collaborative decision-making.”
He added: “Dataiku focuses on integrating technical solutions within a secure and managed corporate environment. Every project is subject to oversight, governance, and the approval of authorised users within a single platform where everyone can work, whether they are programmers, analysts, or business experts. This allows for the delivery of ideas and technologies from the initial prototype to production in days, not months, while saving all work steps and decisions for future reference.”
When asked what distinguishes Dataiku, he said: “The core advantage is that Dataiku is not just a traditional cloud platform; it is designed to be accessible to everyone and to empower organisations to develop customised artificial intelligence solutions according to the nature of their work, not just through ready-made applications from global cloud computing companies. A prominent real-world example is the Swiss company Roche, whose lawyers used Dataiku to build an intelligent legal agent that assists in legal research on patents, to increase work efficiency and save time.”
Governance and trust
Muehmel warned against adopting “absolute trust” in the results of intelligent systems, especially when institutions or governments move to rely entirely on technology that provides answers and decisions with “great confidence” without providing an explanation or logic behind the results. He stressed the need for humans to remain in the decision-making loop, meaning the answer should not be “this is what the artificial intelligence said,” but must be logically justified and subject to review and documentation.
Regarding the levels of these systems and how to apply these tools, he said, “Effective systems begin first by controlling digital access to data and how each category of employees uses the data. This coordination is done through information security teams and technology leaders such as information systems managers, leading to the documentation of all development steps, the application of governance tools, and then monitoring the system after operation to ensure its compliance with regulatory policies and controls.”
He continued, “Among Dataiku’s features is the presence of tools that allow for the monitoring and documentation of all steps and decisions within the platform, which allows technical teams to return to any step or ‘decision point’ and deal with the system in a transparent and auditable manner, even in cases of inquiries from regulatory authorities.”
From experimentation to production
Muehmel explained that the best ways to transition from the initial prototype to production begin with applying artificial intelligence to operations with clear steps, such as responding to technical support tickets or processing specific requests, where a pilot model can be built with the integrated tools within Dataiku in collaboration with work teams. The quality of the results is tested periodically, and the use is gradually expanded while monitoring the quality of performance and feedback, and the possibility of generalising the model if it proves its effectiveness. This is a process that does not stop at the mere press of a button but requires review and successive steps to ensure the appropriateness of the decision and the completion of the application in a safe and effective manner.
Real-world experiences in the region
During the meeting, Sid Bhatia presented a number of application cases in the region within the health, technical support, and retail sectors. He explained that more than 40% of artificial intelligence projects are currently focused on improving health processes, while 30% focus on improving internal procedures such as technical support and analysis, and about 20% for automating business directly.
Bhatia pointed out that the biggest obstacle facing institutions is the absence of effective governance and the lack of clarity of actual use cases, in addition to challenges in data integration and its lack of readiness in the appropriate formats. He pointed out that the solution lies in selecting realistic cases that are easy to measure, test, and highlight their results to managers and operational teams, which helps to build real trust and persuade institutions to expand the scope of the application gradually.
Among the important examples that Bhatia mentioned is the creation of systems to convert calls in support centres to analyse texts and extract common questions, with the building of intelligent agents that suggest the appropriate response and potential paths for the conversation. This application has shown tangible success and high commercial value for one of the local clients.
Dataiku focuses on strengthening growth paths through geographic expansion, product development, and creating strategic partnerships in the field of technology and artificial intelligence. The company allocates huge investments in research and development, as it launched new tools last year in the field of generative artificial intelligence (GenAI) such as Dataiku Answers and Dataiku Stories, which allows institutions to build intelligent applications without the need for programming, and enables them to automate thousands of business processes in a centralised, secure, and managed manner. According to a Forrester study, Dataiku achieves a return on investment of up to 413%, which confirms its commercial feasibility and the speed of its spread among customers.
The company is intensifying its presence in the Middle East by increasing the number of technical support teams and data experts, in addition to partnerships with technology solutions and consulting companies, with the aim of improving customer experiences and driving digital transformation towards higher levels of actual efficiency and dazzle, not just “customer satisfaction.”
Dataiku constitutes a modern model for institutional artificial intelligence that combines technical and human experience and dedicates the philosophy of “managed trust” through governance and transparency. It offers fast, multi-use, and customisable solutions according to the needs of customers in different sectors (health, financial, industrial) with tangible practical experience in the Middle East, Europe, and America. The growth of the platform is accelerating due to the increasing demand for managed artificial intelligence applications that are subject to professional auditing and review.
