In a world where artificial intelligence is often synonymous with text and images, a new approach is emerging that harnesses the power of tabular data. Developed by Professor Devavrat Shah at MIT, this innovative system is designed to facilitate real-time planning and decision-making on a large scale, utilizing structured data formats akin to those found in spreadsheets.
Addressing the Limitations of Traditional AI
While AI systems have proliferated in various sectors, many fall short due to a lack of specific organizational insights. Shah, a principal investigator at MIT’s Laboratory for Information and Decision Systems (LIDS), emphasizes the need for methods that can efficiently manage decision-making processes that occur every second, even with limited computational resources. “With a small amount of resource, you have to do a lot of heavy lifting,” he notes, highlighting the challenge of extracting actionable information from data.
Ikigai Labs and the Foundation Model
In 2019, Shah co-founded Ikigai Labs, which developed a foundation model for tabular, time series data based on extensive research from his lab. This patented model is capable of continuously learning from enterprise data across diverse sources, refining its predictions against actual outcomes. Shah likens the system to graphical models used in GPS technology, which convert sparse satellite data into precise location information.
Real-World Applications and Implications
The primary goal of Ikigai is to enhance forecasting and decision-making capabilities for large enterprises, such as consumer goods manufacturers and pharmaceutical companies. For instance, a consumer electronics firm could utilize this system to predict sales and demand fluctuations based on pricing changes or promotional strategies. Shah explains that the interdependent nature of these processes necessitates a digitized approach to optimize operations.
Integration with Celonis
Recently acquired by Celonis, a leader in operational digitization, Shah now serves as chief scientist while continuing his academic roles. The integration aims to leverage Ikigai’s model to provide companies with tools that can analyze their own data and business processes, facilitating informed decision-making. Shah envisions a future where the digital infrastructure of businesses supports advanced simulations and forecasts, enhancing operational efficiency.
Shah asserts that focusing on structured or time-domain data offers a cost-effective avenue for AI development. “A narrower focus comes with sharper technology,” he states, underscoring the value of this approach in a landscape where many are pursuing broader, less defined goals. By building what he describes as an “enterprise process world model,” Shah aims to fill a critical gap in the current AI landscape.
This article was produced by NeonPulse.today using human and AI-assisted editorial processes, based on publicly available information. Content may be edited for clarity and style.
