Originals · Startup · Startups & Fintech
Arborum Labs raises a pre-seed round to build AI world models that learn how power and telecom networks behave
The Bengaluru startup, founded by IISc professor Himanshu Tyagi and former BCG GenAI lead Pooja Rajdev, is backed by Speciale Invest and Ideaspring Capital and plans a public version of its model within a year.

Arborum Labs, a Bengaluru AI startup, has raised an undisclosed pre-seed round co-led by Speciale Invest and Ideaspring Capital, ETEntrepreneur, Tech in Asia and YourStory report. Angel investors Neeraj Aggarwal, vice-chair for Asia Pacific at BCG, Pranay Mehrotra, managing director and senior partner at BCG, and Sriram Rajamani, corporate vice-president and former managing director of Microsoft Research, joined the round, along with ARTPARK. The company was first incubated at the Foundation for Science Innovation and Development.
The company is building what it calls world models for infrastructure networks. Instead of a digital twin drawn up by hand from engineering specifications, the idea is an AI model that learns cause and effect from a network's own data, so it can forecast demand, optimise day to day operations and flag disruptions before they happen. Its first targets are power and telecom.
Arborum was founded by Himanshu Tyagi, an associate professor at the Indian Institute of Science who works on information theory and machine learning, and Pooja Rajdev, who previously led generative AI work at BCG India. "India needs to leapfrog traditional infrastructure planning and development by directly going to AI for it," Tyagi told ETEntrepreneur, adding that every planning and operating decision on networks like power, telecom and logistics carries high stakes.
The startup says it already works with several nationwide networks in India, including a steel maker on the power side and telecom operators. In a post explaining its investment, Speciale Invest says Arborum works with one telecom operator and two equipment makers, reports about 90% accuracy on hard root cause cases, and has back-tests in power pointing to roughly Rs 25 crore a year in savings for one large consumer. These are the investor's figures and have not been independently verified.
The money will go into product development and commercial expansion. Over the next year, Arborum plans to release a public version of its world model and grow its customer base in India and the US.