Akshat Prakash began research that culminated in Camb.ai’s core technology while he was a computer science student at Carnegie Mellon University, which he parlayed into an artificial intelligence/machine learning engineering role on Apple’s Siri team from 2020-22. By that time, his father, Avneesh, had three decades of experience as a business-builder, spanning sectors including media, entertainment and identity.
Seeking to bridge language divides they saw as underserved by modern technology, the father-son duo launched their AI-powered translation platform in 2023 and have since grown it into an essential localization tool for several major sports properties.
“Traditionally, what usually happens is the father builds a kingdom of a business, and then son takes it over,” said Akshat Prakash, an SBJ New Voices Under 30 honoree in 2024. “We started this together.”
Camb.ai
FOUNDED: 2023
HQ: Dubai / USA
EMPLOYEES: 50
KEY EXECUTIVES: Avneesh Prakash, CEO; Akshat Prakash, CTO
KEY PARTNERS: Imax, MLS, Australian Open, MLSE, EBU
Sports-related use cases of Camb.ai’s production studio include translating live commentary of games, videos-on-demand and — outside of speech — the text of documents. The company was a member of MLS’s inaugural Innovation Lab, in which, for the first time, it simultaneously translated commentary from a live sporting event into multiple languages. Camb.ai also has worked with Tennis Australia, Maple Leaf Sports & Entertainment and the European Broadcasting Union.
It is now bursting into other sectors as well, including a deal with Imax to translate content shown in its theaters.
“Sports content has everything,” Akshat Prakash said. “It has emotion. It has niche lingo. It has background cheering. It has multiple speakers. It’s fast-paced. Anything and everything that could be hard about content exists in sports. Our basic theory was, ‘If we can do sports well, we could do everything well.’ Every vertical we’re working on, you see it because we’re very good at sports.”
Camb.ai delivers its speech-to-speech translations in more than 140 languages through a multipronged process that leverages two AI models developed by the company. The models — a translation engine and a speech synthesizer — consider emotional cues and speech signals.
The end result, in Prakash’s words, is that the translated audio “feels like it’s the same exact emotion, the same exact person,” as the original.
