{"id":21443,"date":"2026-10-19T18:10:29","date_gmt":"2026-10-19T12:40:29","guid":{"rendered":"https:\/\/thankyoubharat.com\/index.php\/2026\/10\/19\/what-indias-190-ai-projects-still-need\/"},"modified":"2026-10-19T18:10:29","modified_gmt":"2026-10-19T12:40:29","slug":"what-indias-190-ai-projects-still-need","status":"publish","type":"post","link":"https:\/\/thankyoubharat.com\/index.php\/2026\/10\/19\/what-indias-190-ai-projects-still-need\/","title":{"rendered":"What India\u2019s 190 AI projects still need"},"content":{"rendered":"<div>\n<p class=\"wp-block-paragraph\"><strong>New Delhi [India], October 19:<\/strong> The government has approved 190 projects under the IndiaAI Mission and committed \u20b910,372 crore to the ecosystem around them. Forty-six sit with startups and MSMEs, and thirty more with early-stage startups. Every one needs recorded Indian material to train on, and most of it has never been collected.<\/p>\n<p class=\"wp-block-paragraph\">The models the mission has already delivered include text-to-speech in Bengali, Gujarati, Kannada, and Malayalam, which is the half of the problem that speaks. Compute was the expensive problem. Data is the slow one.<\/p>\n<h2 class=\"wp-block-heading\">Where the funding meets the recording gap<\/h2>\n<p class=\"wp-block-paragraph\">A model carries the market it was recorded in. Trained on quiet rooms and one language per sentence, it meets India and misses:<\/p>\n<ul class=\"wp-block-list\">\n<li>A support call that names the merchant in English, the installment in Marathi, and the amount in Hindi.<\/li>\n<li>A kirana shelf stacked with regional packaging no global dataset has ever photographed.<\/li>\n<li>A handwritten ledger in a script the model has never been shown.<\/li>\n<li>A pack label in Malayalam, sitting next to one in Hindi.<\/li>\n<\/ul>\n<p class=\"wp-block-paragraph\">Boards are asking for AI results, and technical teams are already building. This is the wall they hit, and the cost is countable: repeat calls, escalations to a human, and a pilot that never leaves the demo.<\/p>\n<h2 class=\"wp-block-heading\">On request, not off the shelf<\/h2>\n<p class=\"wp-block-paragraph\">Two things trade under the one name:<\/p>\n<ul class=\"wp-block-list\">\n<li><strong>Off the shelf.<\/strong> A finished file, assembled once and sold to whoever buys it next.<\/li>\n<li><strong>On request.<\/strong> Recorded in a written brief, in named districts, by people paid to supply it.<\/li>\n<\/ul>\n<p class=\"wp-block-paragraph\">Off-the-shelf files were assembled for someone else\u2019s brief. Projects that need named districts and languages need the second kind.<\/p>\n<p class=\"wp-block-paragraph\">That is the work Rwazi does. Rwazi collects <a href=\"https:\/\/rwazi.com\/ai-datasets?utm_source=business-today&amp;utm_medium=referral&amp;utm_campaign=india-pr-ai-datasets&amp;utm_content=ai-datasets\" target=\"_blank\" rel=\"noopener\">real-world AI training data<\/a> on request, project by project, through a network of real people across 190+ countries who agreed to share what they record. The same network is how an Indian team reaches a market outside India without rebuilding collection for it.<\/p>\n<figure class=\"wp-block-image size-large\"><img fetchpriority=\"high\" fetchpriority=\"high\" decoding=\"async\" width=\"1024\" height=\"576\" src=\"https:\/\/thankyoubharat.com\/wp-content\/uploads\/2026\/10\/PNN-2026-10-19T181818325-1024x576-1.jpg\" alt=\"\" class=\"wp-image-89988\" srcset=\"https:\/\/thankyoubharat.com\/wp-content\/uploads\/2026\/10\/PNN-2026-10-19T181818325-1024x576-1.jpg 1024w, https:\/\/pnndigital.com\/wp-content\/uploads\/2026\/10\/PNN-2026-10-19T181818.325-150x84.jpg 150w, https:\/\/pnndigital.com\/wp-content\/uploads\/2026\/10\/PNN-2026-10-19T181818.325-300x169.jpg 300w, https:\/\/pnndigital.com\/wp-content\/uploads\/2026\/10\/PNN-2026-10-19T181818.325-768x432.jpg 768w, https:\/\/pnndigital.com\/wp-content\/uploads\/2026\/10\/PNN-2026-10-19T181818.325-450x253.jpg 450w, https:\/\/pnndigital.com\/wp-content\/uploads\/2026\/10\/PNN-2026-10-19T181818.325.jpg 1200w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n<h2 class=\"wp-block-heading\">All nine types, collected on request<\/h2>\n<p class=\"wp-block-paragraph\">One Indian project usually crosses several of these, and a single brief can carry all nine.<\/p>\n<ol class=\"wp-block-list\">\n<li><strong>Audio and speech.<\/strong> Support calls and read speech with transcripts. Text-to-speech Hindi is solved; a voice bot that listens across accents and code-mixed sentences still needs a speech dataset recorded here.<\/li>\n<li><strong>Video.<\/strong> A video dataset shot rather than scraped, including first-person capture for robotics.<\/li>\n<li><strong>Image.<\/strong> An image dataset captured on request. Most computer vision projects in India stall here: packaging, signage, storefronts, and handwriting.<\/li>\n<li><strong>Text.<\/strong> Written in the Indian language in the script people actually use, mixed-script included.<\/li>\n<li><strong>Mobile sensor.<\/strong> Location, motion, and device signals contributed from handsets at district resolution.<\/li>\n<li><strong>Multimodal data.<\/strong> Two or more recorded together and aligned, for models that read context.<\/li>\n<li><strong>Consumer data.<\/strong> What people bought, where, and what they chose instead, below the metro boundary.<\/li>\n<li><strong>Retail shelf.<\/strong> Shelf photography from real stores for pricing, availability, and planogram work.<\/li>\n<li><strong>Medical.<\/strong> Packaging, labels, and pack imagery, where medical AI in India runs shortest.<\/li>\n<\/ol>\n<h2 class=\"wp-block-heading\">Send a brief, and get recorded data back<\/h2>\n<p class=\"wp-block-paragraph\">Before any contract is signed:<\/p>\n<ul class=\"wp-block-list\">\n<li><strong>What to send.<\/strong> Type, languages, districts, volume, and date. Or the single line that matters most: what the model keeps getting wrong and where.<\/li>\n<li><strong>What comes back.<\/strong> A sample recorded by contributors in the markets the brief names, not an extract from a catalogue.<\/li>\n<li><strong>What it costs.<\/strong> No fee for the sample, and no agreement is signed first.<\/li>\n<li><strong>Where it goes.<\/strong> <a href=\"https:\/\/rwazi.com\/ai-datasets\/contact?utm_source=business-today&amp;utm_medium=referral&amp;utm_campaign=india-pr-ai-datasets&amp;utm_content=contact\" target=\"_blank\" rel=\"noopener\">Rwazi\u2019s AI datasets team<\/a>.<\/li>\n<\/ul>\n<h2 class=\"wp-block-heading\">How the data collection runs<\/h2>\n<ol class=\"wp-block-list\">\n<li><strong>Agreement.<\/strong> A master service agreement and a statement of work lock the brief.<\/li>\n<li><strong>Samples.<\/strong> Recorded against it and checked before anything scales.<\/li>\n<li><strong>Scale.<\/strong> Raw phone capture and studio-grade capture run on the same rails, so one project can ask for both.<\/li>\n<\/ol>\n<p class=\"wp-block-paragraph\">Nothing is inferred. Every recording is supplied by the person who made it.<\/p>\n<p class=\"wp-block-paragraph\">\u201cA model that fails in a district town fails for everyone who lives there,\u201d said Joseph Rutakangwa, CEO, Rwazi. \u201cThe recording has to happen where the customer is, in the language the customer uses.\u201d<\/p>\n<h2 class=\"wp-block-heading\">What this decides<\/h2>\n<p class=\"wp-block-paragraph\">India\u2019s AI money is already committed. The 190 approved projects will be judged on whether their models work in the districts their customers live in, rather than in a demonstration, and that comes down to what was recorded, in which languages, and under which conditions. The computer arrived first. The data decides what it produces.<\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>New Delhi [India], October 19: The government has approved 190 projects under the IndiaAI Mission and committed \u20b910,372 crore to the ecosystem around them. Forty-six sit with startups and MSMEs, and thirty more with early-stage startups. Every one needs recorded Indian material to train on, and most of it has never been collected. The models<\/p>\n","protected":false},"author":1,"featured_media":21444,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[13],"class_list":["post-21443","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-business","tag-business"],"_links":{"self":[{"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/posts\/21443","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/comments?post=21443"}],"version-history":[{"count":0,"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/posts\/21443\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/media\/21444"}],"wp:attachment":[{"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/media?parent=21443"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/categories?post=21443"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/thankyoubharat.com\/index.php\/wp-json\/wp\/v2\/tags?post=21443"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}