If your target market has 22 official languages and its people speak in over 19,000 dialects, does it make sense to offer a text-only AI chatbot that can function best in a couple languages?
That’s the question Indian AI startup Sarvam has been working to solve, and on Tuesday it launched a series of offerings, including a voice-enabled AI bot that supports more than 10 Indian languages, betting that people in the country would prefer to talk to an AI model in their own language rather than chat with it over text. The startup is also launching a small language model, an AI tool for lawyers, as well as an audio-language model.
“People prefer to speak in their own language. It’s extremely challenging to type in Indian languages today,” Vivek Raghavan, co-founder of Sarvam AI, told TechCrunch.
The Bengaluru-based startup, which primarily targets businesses and enterprises, is pitching its AI voice-enabled bots for a number of industries, particularly those relying on customer support. As an example, it pointed to one of its customers: Sri Mandir, a startup that offers religious content, has been using Sarvam’s AI agent to accept payments, and has processed more than 270,000 transactions so far.
The company said its AI voice agents can be deployed on WhatsApp, within an app, and can even work with traditional voice calls.
Backed by Peak XV and Lightspeed, Sarvam plans to price its AI agents starting at ₹1 (approximately 1 cent) per minute of usage.
The startup is building its voice-enabled AI agents on top of a foundational, small language model, called Sarvam 2B, that’s trained on a data set of 4 trillion tokens. The model is completely trained on synthetic data, according to Raghavan.
AI experts often advise caution when using synthetic data — essentially data generated by a large language model that aims to replicate real-world data — to train other AI models, because LLMs tend to hallucinate and make up information that may not be accurate. Training AI models on such data may serve to exacerbate such inaccuracies.
Raghavan said Sarvam opted to use synthetic data due to the extremely limited availability of Indian language content on the open web. The startup has developed models to clean and improve the data first used to generate the synthetic datasets, he added.
The founder claimed that Sarvam 2B will cost a tenth of anything comparable in the industry. The startup is open-sourcing the model, hoping that community will further build upon it.
“While the large language foundational models are very exciting, you can achieve an experience that is superior, more specific, lower-cost and with reduced latency using small language models,” Raghavan said. “If you want to perform a query or two in a week or a month, you should use the large language models. But for use cases requiring millions of daily interactions, I believe smaller models are more suitable.”
The startup is also launching an audio-language model, called Shuka, built on its Saaras v1 audio decoder and Meta’s Llama3-8B Instruct. This model is also being open-sourced, so developers can use the startup’s translation, TTS, and other modules to build voice interfaces.
And, there’s another product dubbed “A1” — a generative AI workbench designed for lawyers that can look up regulations, draft documents, redact them and extract data.
Sarvam is one of the small group of Indian startups advocating for use cases that align with the country’s interests and contribute to the government’s efforts to develop its own bespoke AI infrastructure.
Governments across the world are increasingly pursuing “sovereign AI” – AI infra that’s developed and controlled at the national level. The purported aim of such efforts is to safeguard data privacy, stimulate economic growth and tailor AI development to their cultural contexts. The United States and China currently have the biggest investments in this space, and India is following with its “IndiaAI” program and language-specific models.
One of the initiatives under the IndiaAI program is called IndiaAI Compute Capacity, and the plan is to establish a supercomputer powered by at least 10,000 GPUs. One of the models being developed, dubbed Bhashini, aims to democratize access to digital services across various Indian languages.
Raghavan said his startup is ready to contribute to the IndiaAI program. “If the opportunity arises, we will work with the government,” he said in the interview.
Sarvam’s voice-enabled AI agents can be deployed on WhatsApp, within an app, and can even work with traditional voice calls.
The irony was not lost on her. Growing up the daughter of a family obsessed with car racing, Danielle Walsh had become — in her late 20s — the head of…
Opera is releasing its redesigned Opera One browser on iOS as a stable release after testing it in the beta phase for weeks. The new browser has a bottom placed…
In Puerto Rico, tax breaks enacted in 2012 aimed to juice the economy by encouraging mainland U.S. citizens to do business and live on the island, where they could apply…
Elon Musk and Donald Trump’s joint X Spaces event appears to have crashed Monday afternoon. The conversation between the owner of X and the former President was scheduled for 5…
Antler, the Singapore VC that focuses on early-stage investments, just closed its second Southeast Asia fund. It’s raised $72 million to double down on startups in Singapore, Indonesia, Vietnam and…
It racked up around 18,000 users, made 8,000 matches, and gathered a lot of insights on the current dating scene.
Fram2 would launch into a polar orbit from Florida in late 2024, after which it will stay up at 425-450 kilometers of altitude for three to five days.
A class action lawsuit filed by artists who allege that Stability, Runway and DeviantArt illegally trained their AIs on copyrighted works can move forward, but only in part, the presiding…
Tally, a nine-year-old fintech that helped consumers manage and pay off their credit card debt, has shut down, according to the company. In a LinkedIn post that was shared earlier…
Dawn Aerospace Mk-II is essentially “an aircraft with the performance of a rocket, not a rocket with wings.”
The U.S. Securities and Exchange Commission (SEC) is suing a crypto startup, NovaTech, for allegedly fraudulently raising more than $650 million from over 200,000 investors, many in the Haitian-American community.…
The FBI’s takedown of the Radar/Dispossessor ransomware and extortion gang is a rare win in the fight against ransomware.
Featured Article
Some of the largest, most damaging breaches of 2024 already account for over a billion stolen records. Plus, some special shout-outs.
16 hours agoIn the last 12 months, Balderton has announced 12 new investments.
TikTok looks to be taking on popular messaging services like Meta’s WhatsApp and Apple’s Messages, as the company announced on Monday that it’s adding group chats to its platform. You…
There’s a fascinating look by John Herrman over at NYMag today at one of the big proposed uses of AI: summarizing content. We all need things summarized, right? Everybody’s too…
Waymo plans to start testing its fully autonomous vehicles with no human safety driver on freeways in the San Francisco Bay Area this week. Its employees will be the first…
Anduril and Palantir delivered the first Tactical Intelligence Targeting Access Node (TITAN) — the first major milestone in its $178 million contract.
Google Pixel 8 devices made in India start rolling off the production lines just ahead of the Pixel 9 launch.
Apple has threatened to remove creator platform Patreon from the App Store if creators use unsupported third-party billing options or disable transactions on iOS, instead of using Apple’s own in-app…
Elevate your brand’s presence at TechCrunch Disrupt 2024 in San Francisco by hosting a custom Side Event during “Disrupt Week,” taking place October 26 through November 1. Engage face-to-face with…
Meta and Universal Music Group (UMG) announced on Monday the expansion of their multi-year music licensing agreement, which enables users to share songs from UMG’s music library across Meta’s platforms…
WeRide, a Chinese autonomous vehicle company, is officially gearing up for a U.S. public debut, over a year after China started easing its effective ban of foreign IPOs. The company is…
When users click on an event on Polymarket, they will now see a summary of news related to the event based on search results from Perplexity.
The U.K. antitrust regulator has confirmed that it’s carrying out an early-stage inquiry into Synopsys‘ plans to buy Ansys. The Competition and Markets Authority (CMA) has opened an “invitation to…
Here is a look back at the top security research from the annual hacker conferences, Black Hat and Def Con 2024.
Cross-border payments for businesses in emerging markets remain significantly untapped, despite small to large businesses using banks and legacy fintechs to transact trillions of dollars in transaction volume annually. A…
BT, the U.K.’s former incumbent telecoms carrier, is picking up a major new investor today as telecoms companies look for stronger footing in the rapidly shifting technology and communications market.…
X, the social media platform owned by Elon Musk, has been targeted with a series of privacy complaints after it helped itself to the data of users in the European…