This Startup Lets You Skip Awkward Movie Moments with AI Control - CHOICE Startup Review
Choice
Somewhere between a censor board's rulebook and a recommendation algorithm's guesswork lies a problem nobody in media has fully solved: knowing, at scale, exactly what's inside the content you're streaming. A single OTT platform can carry tens of thousands of hours of video across a dozen languages, and every hour of it needs to be checked for nudity, profanity, and violence, tagged with the right metadata, dubbed or subtitled for a new market, and served to the right viewer at the right moment. Do that manually, and you're either slow, inconsistent, or both.
Choice AI is betting that the entire of moderation, localization, personalization, and compliance shouldn't live in five different vendor contracts. It's built a single AI layer that platforms, broadcasters, and regulators can plug into, and it's doing this at a moment when India's media and entertainment sector is scaling faster in regional languages than in Hindi. Every frame, in Choice AI's pitch, gets smarter, safer, and more personal without a human watching all of it first.
What makes the story more interesting is who's behind it. Choice AI isn't a lone founder's side project; it's built inside PanScience Innovations (PSI), one of India's larger deep-tech and AI venture studios, which co-founds and operates ventures across healthtech, fintech, legaltech, and media rather than simply funding them from the outside.
What is Choice AI?
Choice AI is a content intelligence platform built to help media companies moderate, localize, personalize, and optimize video, audio, and text content in over 30 languages. Its engine, marketed under the hood as NovaCast OS, handles the unglamorous but expensive parts of running a media business: scene-level nudity and profanity detection, violence and gore flagging, automated metadata tagging, age-rating and compliance checks, and rights and territory management, all wired into plug-and-play APIs so OTT, D2H, and broadcast teams can deploy without rebuilding their stack.
Layered on top of that compliance backbone is a more consumer-facing ambition. Choice AI's tools already sit alongside major Indian and global platforms, including Zee5, Hotstar, SonyLIV, Jio Cinema, MX Player, Shemaroo, Amazon Prime, and YouTube, and the company is pushing this beyond the browser into smart TVs, set-top boxes, and mobile. The idea isn't just to filter content for platforms, but to eventually let individual viewers mute topics instead of just people: skipping or watching specific tagged scenes, or building their own custom categories to follow, rather than relying on a platform's one-size-fits-all rating.
That dual identity, a B2B compliance and localization engine on one side, a self-moderation layer for audiences on the other, is also why Choice AI has found itself in rooms it wouldn't otherwise be in. It's been selected for the MediaTech Acceleration Program run by IIML EIC and Startup India, is part of the Microsoft for Startups Founders Hub, and has been in active conversation with India's Central Board of Film Certification and other ministries on what proactive, user-led content moderation could look like as an alternative to blanket regulation.
Backed by PanScience Innovations
Choice AI's positioning starts to make more sense once you place it inside PanScience Innovations' portfolio. PSI operates as a venture studio rather than a traditional accelerator or fund, it co-builds ventures internally, sharing engineering, GTM, and operational talent across its portfolio companies instead of leaving each one to hire and figure things out from zero. Choice AI is one of several ventures under this model, sitting specifically in what PSI frames as India's "media and content intelligence" category: a category the studio argues is underserved relative to the size of India's media sector and the speed at which regional-language content is overtaking Hindi programming.
That studio backing shows up in how Choice AI talks about itself: less like a scrappy point-solution chasing one use case, and more like infrastructure meant to sit underneath an entire content pipeline: from ingest, to compliance, to the moment a viewer presses play.
