Geostar develops AI search in over 15 languages in 8 weeks
Original headline지오스타, AI 검색 15개 넘는 언어로 8주 만에 개발
Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.
JioStar built conversational video and sports search for JioHotstar in more than 15 languages in roughly eight weeks, letting users query movies, series, and specific cricket moments and player records by voice or text. The feature was built on OpenAI APIs, with senior architect Vijay Seshadri telling OpenAI that repeated evaluation and verification of search answers was central to the process. Seshadri said the team had to balance accuracy, response latency, and cost from the start of development. The rollout matters because it shows how quickly a large Indian streaming platform can localize AI-driven search across a linguistically diverse market using third-party models. What to watch next is how JioStar tunes that accuracy-latency-cost trade-off as usage scales and whether the evaluation framework holds across additional languages and content types.
Muhabbit's AI wrote this summary in its own words from TokenPost's report and translated it from Korean. It can get details wrong, so check the source.
Named in this story
JioStar, JioHotstar, Vijay Seshadri, OpenAI