Ayushmann Khurrana, Sharvari’s 'Yeh Prem Mol Liya' to hit theatres on Nov 27Ayushmann Khurrana, Sharvari’s 'Yeh Prem Mol Liya' to hit theatres on Nov 27 Apr 21, 2026, 12:30:17 IST PTI Share via Copy link Mumbai, Filmmaker Sooraj Barjatya's next feature "Yeh Prem Mol Liya"
Iran’s top negotiator and parliament speaker, Mohammad Bagher Ghalibaf, has accused US President Donald Trump of attempting to pressure Tehran into submission, warning that Iran is prepared to escalate if negotiations are conducted “under the shadow of threats
Swedish private equity giant EQT has successfully raised $15.6 billion for its Asia-focused buyout fund, EQT BPEA IX, exceeding its target. This oversubscribed fund, backed by a diverse global investor base, will target leading companies in technology, healthcare, and industrial sectors
, the parent of online stock trading platform , has acquired algo-based investing and trading startup in a cash-and-stock deal. The deal, according to sources, is valued at $5–6 million. ET had first on November 19, 2025.Startzy is estimated to be clocking an annual revenue run rate of Rs 7-8
When a person gets an ECG, a stress test, or a blood test done, a reassuring “normal" report feels like a green signal. But modern cardiology is increasingly challenging this idea, pointing out that normal test results do not always mean a healthy heart.When a person gets an ECG, a stress test
A Canadian woman was shot dead and 13 others were hurt in a shooting Monday (April 20, 2026) at Mexico's famed Teotihuacan archaeological site, authorities said. The gunman, whose nationality was unknown, killed himself after opening fire at the popular tourist destination in central Mexico
There was a time when indulgence came with a caveat, earned through restraint, justified as a “cheat day," and often followed by a quiet sense of guilt. That narrative is steadily evolving. Across urban India, food is no longer framed as a trade-off between pleasure and wellness. Instead
A passenger at Pune airport claimed he was not allowed to board his flight to Bengaluru despite arriving on time. He said he had a confirmed ticket and reached the airport well within the check-in window. The issue, he was told, was that the flight had been overbooked
Tulu, Bodo, Kashmiri: Startups are teaching AI models Indian dialects
This article was originally published in Rest of World, which covers technology's impact outside the West.
When Amrith Shenava began experimenting with large language models shortly after the launch of ChatGPT, he quickly realized that Tulu – the language he and some 2 million people spoke in the southern Indian state of Karnataka – had virtually no digital data set. He decided to build one.
Shenava, who has a degree in computer science from Kent State University in Ohio, had earlier launched a translation app, and a language learning app for Tulu. To build the data set for the LLM, he had to collect voice and text data from native speakers including teachers, professionals, homemakers, and members of the Tulu diaspora.
“Most AI systems are built in the US. They don't understand Indian languages or contexts,” Shenava, the 27-year-old founder of TuluAI, told Rest of World. “We need our own models that represent us.”
India has more than 1,600 languages and dialects, but most artificial intelligence systems cater to those that are widely spoken. OpenAI's ChatGPT supports more than a dozen Indian languages including Hindi, Tamil, and Kannada, the dominant language in Karnataka. Google's Gemini can chat with users in nine Indian languages.
Spurred by their success, and keen to be a part of the rapid global transition to AI, a handful of Indian startups are building AI tools for so-called low-resource languages such as Tulu, Bodo, and Kashmiri, which have a limited online presence and few written records. The startups are having to build data sets nearly from scratch.
TuluAI holds storytelling sessions and workshops in rural areas, in which local residents – particularly women and elders – narrate their stories, or are asked to read texts and simulate everyday conversations. Participants are taught to record and label the data. Each workshop of one to two days produces over 150 hours of labeled voice and text data, Shenava said.
The startup also collects WhatsApp voice notes from anyone who wishes to send one, with annotators checking transcripts and labels for accuracy.
“Major translation tools miss the context that gives meaning to words. The only way to fix that is to use authentic, human-recorded data that reflects real-life language use,” Shenava said. “The goal is for the model to talk like a native speaker. We want it to understand humor, idioms, and cultural context. So we're building slowly, verifying every sample.”
Across the country, in the northeastern state of Assam, Kabyanil Talukdar, the 25-year-old co-founder of Aakhor AI, follows a similar process to build data sets in Bodo and Assamese. Talukdar's team conducts community workshops and classes, and holds voice-note drives via WhatsApp groups, with simple daily prompts like “Talk about your morning tea.”
Each submission is tagged with metadata such as dialect, region, and speaker demographics to ensure diversity. The clips, 20-60 seconds long, are processed, transcribed, and anonymised. Each three-month campaign produces over 5,000 voice samples, Talukdar told Rest of World.
“When people see that their voices help preserve their language, they feel ownership,” he said. “They are driven by the shared goal of creating AI that understands and speaks their native language.”
Big tech LLMs such as GPT and Meta's Llama are trained on a wide range of data, including in languages other than English. Yet their performance in low-resource languages can be unpredictable, particularly in dialects and local idioms. Countries keen to support their languages and become self-sufficient in AI are building their own multilingual LLMs, which can support translation, speech recognition, and tools for customer service, education, health care, and other applications.
These include the Chile-led LatamGPT project, Southeast Asia's Sealion, and efforts by Masakhane – a grassroots organisation that aims to build AI data sets and tools in African languages. India's BharatGPT and Sarvam support many major Indian languages, and the government is building open-source models for several languages under the Bhashini project.
It is not easy.
Tulu's ancient script lacks a Unicode standard that would allow computational processing of text. Shenava's team is digitising literature written in the script, and training the model to identify patterns. While more complicated, the process helps capture the cultural nuance that is often lost in translation, he said.
The team avoids AI-generated or machine-translated data, which is often riddled with grammatical errors, made-up words and phrases, and other inaccuracies, he said.
“Even open-source models produce text that doesn't make sense. That's why we decided to build it from scratch,” Shenava said. This also ensures ethical data use, he said. “We don't use any personal data without explicit permission.”
Aakhor AI's models are voice-first, targeting areas with low literacy and weak internet access. The company recruits speakers from underrepresented areas to prevent dominant dialects from overshadowing smaller ones, and ensure “balanced sampling,” Talukdar said.
For Saqlain Yousef, it was the fear that Kashmiri – a language spoken by about 7 million people in India – might disappear that drove him to build the KashmiriGPT app using OpenAI's application programming interface.
The platform accepts input in English as well as Kashmiri written in the Roman script, and generates responses in the Kashmiri script, Roman Kashmiri script, and English.
“Our language is vulnerable and at risk of disappearing. So I took matters into my own hands,” the 25-year-old told Rest of World. “This will help preserve Kashmiri in the AI age.”
Yousef is right to be concerned, C Vanlalawmpuia, an independent researcher in language and AI, told Rest of World.
“These languages are already marginalised, and without proper digital representation, they risk disappearing from online spaces entirely,” he said.
AI makes it easier to preserve a language through translation tools, transcription systems, and data sets that can make a language more visible and accessible, according to Vanlalawmpuia. But the lack of digital resources and funding are a challenge, and community-led efforts are one way to sustain the platforms, he said.
AI platforms from deep-pocketed big tech firms including OpenAI, Google, and Perplexity are also targeting India. The country is already the biggest market for ChatGPT outside the US, and OpenAI this month offered its ChatGPT Go service free for a year to users in India.
Aakhor AI is aware of its challenge. “We don't compete with GPT on scale,” Talukdar said. “We compete on relevance.”
By sourcing data from the ground, the community is involved in preserving linguistic diversity and advancing linguistic inclusion, Shenava said.
“Anyone can contribute. That's how language preservation will happen,” he said. “If AI can help keep it alive, that's worth all the effort.”
For Rita D'Souza, a 32-year-old primary schoolteacher in coastal Karnataka, TuluAI is already making a difference, helping students improve their pronunciation and spelling, she told Rest of World.
Tauseef Ahmad is a freelance journalist based in Delhi.
Sajid Raina is a freelance journalist based in Delhi.
This article was originally published in Rest of World, which covers technology's impact outside the West.
Source: Scroll
Related Posts: Bondi Beach gunman originally from India DIIs continue to provide strong support to markets driven by strong domestic X to award $1 million to top article in push for long-form writing Explained | What inclusion in Article 240 would mean for Chandigarh Modi Govt's Article 370 Row Erupts In Punjab As Centre Moves To Bring Chandigarh Under Article 240 Removing Article 370 was a true tribute to Sardar Patel Modi Govt's Article 370 Netanyahu shares article claiming Epstein worked for Israeli intelligence Final voter list to be published on February 21
Iran has executed Amirali Mirjafari. He was accused of leading a network linked to Israel's intelligence service. Mirjafari was also convicted of setting fire to a mosque in Tehran during recent protests. The Supreme Court upheld his death sentence. The execution took place early Tuesday
1 hours ago
India's security agencies gain a new edge with the indigenously developed satellite imaging system Prajna. This AI-enabled technology will bolster internal security and counter-terrorism efforts. Meanwhile, the Indian Navy successfully tested the Air Droppable Container ADC-150
1 hours ago
Lamine Yamal in awe of Lionel Messi's omnipresent legacy: ‘He’s not a legend only in his sports'Lamine Yamal praised Lionel Messi as the best player in history during a recent event. The Barcelona teenager emphasised Messi's influence beyond football. Updated on: Apr 21
1 hours ago
An image of an Israeli soldier apparently hitting a statue of Jesus with a sledgehammer in southern Lebanon has prompted widespread condemnation after being widely shared online. Israel's Prime Minister Benjamin Netanyahu said he was "stunned and saddened" by what happened. Foreign Minister
1 hours ago
An Air Arabia flight bound for Thiruvananthapuram International Airport from Abu Dhabi was suspected to have experienced a bird strike during takeoff on Tuesday (April 21, 2026). Despite the suspected bird hit, the crew continued the flight in accordance with standard safety procedures
1 hours ago
From Steve Jobs to Tim Cook: How Apple’s top leadership quietly changed over 15 yearsTim Cook has announced his retirement as Apple’s CEO. Here’s a look at key executives who exited Apple since the Steve Jobs era and where they are now. Updated on: Apr 21
1 hours ago
For the longest time, my microwave was only used to reheat food. It did the job, but that was about it. Every time I thought about baking a cake or grilling something at home, I either dropped the idea or depended on another appliance. If you’ve had a similar experience, you’re not alone
1 hours ago
Eleven persons were injured in a clash at a gurdwara in Germany on Sunday, India Today reported. More than 40 persons were involved in the altercation in Moers town, ANI reported. The incident led the police to intervene, German newspaper Neue Ruhr Zeitung reported
1 hours ago
At least 11 individuals were wounded after a violent altercation broke out at a gurdwara in the Duisburg area of Moers in Germany. Clips circulating online, reportedly filmed inside the place of worship, show members of two groups engaging in a fierce clash
1 hours ago
An Indian-origin woman based in the United States has sparked a massive debate after explaining why many Indians who move abroad, especially to America, often do not return to India even after years of working there. In a video circulating widely on social media, the woman
1 hours ago
India, like many other developing countries, has been in the midst of a learning crisis, as evidenced by Annual Status of Education Reports (ASER), notwithstanding the marginal improvements seen in the last few years. Yet, the absence of learning outcomes
1 hours ago
Mobile Ban At Kedarnath Dham: Devotees To Face Legal Action For Photos, Videos Or Reels Inside Temple In a strict move to preserve the sanctity of Kedarnath Temple, authorities have imposed a complete ban on mobile phone usage within the temple premises
1 hours ago
Kurseong: Union Home Minister Amit Shah on Tuesday said the decades-old Gorkha issue will be resolved within six months if the BJP forms the government in West Bengal. He said no party other than the BJP could find a solution acceptable to the Gorkhas
1 hours ago
Editorial independence is core to our work. Some links may earn us a commission, without influencing our opinions.From 9 AM meetings to 9 PM plans: These 5 V-neck kurtis do it all effortlesslyIf your schedule rarely allows outfit changes, V-neck kurtis are a smart wardrobe investment
1 hours ago
There’s something oddly moving about looking back at a time when cinema didn’t exist in India… no stars, no studios, no Friday releases. Just an idea waiting to happen. Over a century later, as Indian films dominate screens across the world, it all traces back to one quiet but decisive moment
1 hours ago
4 injured in clash at gurdwara in Germany; police launch probe4 injured in clash at gurdwara in Germany; police launch probe Published on: Apr 21, 2026 12:37 PM IST PTI Share via Copy link London, Four people were injured after a clash broke out between two groups at a gurdwara in Germany's Moers
1 hours ago
South Korean companies , and on Tuesday announced the launch of the , a Rs 6,000-crore vehicle aimed at investing in high-growth .The announcement was made on the sidelines of South Korean President Lee Jae Myung’s visit to India, during which several memorandums of understanding (MoUs) were
1 hours ago
On April 16, the Indian government introduced three Bills in the Lok Sabha: the Constitution (131st Amendment) Bill, the Delimitation Bill, and the Union Territories Laws (Amendment) Bill. The proposed laws would expand the size of the Lok Sabha from 543 seats to 850 seats
1 hours ago
United States President Donald Trump on Monday said that his country will not lift its blockade on Iranian ports until Tehran agrees to a peace deal to end the war in West Asia. “The blockade, which we will not take off until there is a ‘deal,’ is absolutely destroying Iran
1 hours ago
Apple has named John Ternus as its new chief executive to replace Tim Cook who is stepping down after 15 years of leading the technology giant. Ternus, the current head of hardware engineering who has been at Apple for 25 years, will take over on 1 September and Cook will become executive chairman
1 hours ago
The Congress on Monday demanded an investigation into the latest killings in Manipur, including those of two children. The Opposition party called for an immediate and time-bound probe and demanded urgent measures to restore law and order in Manipur
1 hours ago
India’s output of eight core infrastructure sectors slipped into contraction in March, falling 0.4 per cent year-on-year as weakness in coal, crude oil, fertilisers and electricity offset gains in steel, cement and natural gas. This is against the 2.8 per cent expansion recorded in February
1 hours ago
Ravens unveil bold ‘Next Flight’ look ahead of new NFL season; praised by Lamar JacksonThe Ravens introduced their 'New Flight' uniform with a matte-black helmet, and earned praise from players including Lamar Jackson and Derrick Henry. Published on: Apr 21
1 hours ago
JKBOSE 10th Result 2026 declared at jkbose.nic.in, direct link to check Class 10 results hereJKBOSE 10th Result 2026 has been declared. The direct link to check Class 10 results is given here. Published on: Apr 21, 2026 12:28 PM IST By HT Education Desk | Edited by Papri Chanda Share via Copy link
1 hours ago
US actor and producer Patrick Muldoon, known for his roles in the famous soap Days of Our Lives and the movie Starship Troopers, has died at the age of 57, according to US media reports. Muldoon, who also starred in another hit 1990s US TV show, Melrose Place
1 hours ago
Neurologist explains how devouring jalebis left 33-year-old man paralysed; shares diagnostic case studySometimes the effect of sweet indulgence highlights underlying health concerns. Dr Kumar shares one such case study. Published on: Apr 21, 2026 12:25 PM IST By Debapriya Bhattacharya Share via
1 hours ago
Prime Minister Narendra Modi on Saturday launched a sharp attack on opposition parties after the Women’s Reservation Bill failed to secure the required majority in the Lok Sabha, accusing them of sabotaging a key reform for women. Addressing the nation
1 hours ago
