Helping AI models to meet the real world
Through research and entrepreneurship, Professor Devavrat Shah is helping to design methods that can handle constant decision-making using limited computational resources.
Artificial intelligence (AI) models are increasingly being used by businesses to improve forecasting, planning, and decision-making, but many lack the detailed information necessary for effective use. Devavrat Shah, a researcher at MIT's Laboratory for Information and Decision Systems, has been working on designing AI methods that can handle real-time decision-making with limited computational resources.
Shah's foundation model for tabular, time series data, developed at Ikigai Labs, can learn from enterprise data at scale and continuously improve predictions by testing them against real outcomes. This system extends the concept of graphical models used in GPS devices and digital watches to handle structured data like spreadsheets.
Companies like consumer goods manufacturers and pharmaceutical firms could benefit from Ikigai's forecasting and decision-making technology. Shah's goal is to help Celonis, the company that acquired Ikigai, integrate the system with companies' existing data and processes to provide real-world analyses and improve business operations.
By focusing on structured or time-domain data, Ikigai aims to provide a cost-effective version of AI that goes beyond the current buzz around building "world models" for various industries.
Written by urgent.news from MIT News AI's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
Also reported by 1 other outlet
- Helping AI models to meet the real world news.mit.edu
