Sarvam AI Launches Groundbreaking GenAI Platform for India

AI Agents at INR 1 Per Min Could Really Help Scale AI Adoption in India

Are AI agents the next big thing? The co-founders of Sarvam AI definitely think so. One of the startup's theses is that consumers of AI will use generative AI models not just as a chatbot, but to perform tasks and achieve goals and that too, through a voice interface rather than text. At an event held in Bengaluru on August 13th, Sarvam AI announced Sarvam Agents. While the startup, which is backed by Lightspeed, Peak XV, and Khosla Ventures, is not the only company building AI agents, what stood out was the pricing. The cost of these agents starts at just one rupee per minute. According to co-founder Vivek Raghavan, enterprises can integrate these agents into their workflow without much hassle. "These are going to be voice-based, multilingual agents designed to solve specific business problems. They will be available in three channels - telephony, WhatsApp, or inside an app," Raghavan told AIM in an interaction prior to the event. These agents could be integrated into contact centres and various applications across multiple industries, including insurance, food and grocery delivery, e-commerce, ride-hailing services, and even banking and payment apps. For example, they could streamline customer service operations in insurance by handling policy inquiries, make reservations, assist with financial transactions, facilitate order tracking and customer support in food delivery, and manage ride requests and driver communications in ride-hailing apps. A technology that offers this capability at just a rupee per minute could be transformative. AI adoption could see substantial growth with AI agents, and Sarvam AI's mission is to make this a reality. Meta, which owns WhatsApp and other major social media platforms like Facebook and Instagram, introduced Meta AI to all these platforms. Meta AI can be summoned in group chats for planning and suggestions, it can make restaurant recommendations, trip planning assistance, and also provide general information. However, Sarvam AI claims their generative AI stack could help AI scale in India compared to others. Their models perform better in Indic languages than the Llama models, which is powering Meta AI. During the event, the startup demoed their models, which managed to outperform certain models in Indic language tasks. The startup is currently making its agents available in Hindi, Tamil, Telugu, Malayalam, Punjabi, Odia, Gujarati, Marathi, Kannada, and Bengali, and plans to add more languages soon. Interestingly, given the backgrounds of the co-founders, especially Raghavan, who has helped Aadhaar scale significantly in India, the startup is well-positioned to drive widespread AI adoption and impact. Raghavan served as the chief product officer at the Unique Identification Authority of India (UIDAI) for over nine years. As of September 29, 2023, over 138.08 crore Aadhaar numbers were issued to the residents of India. As part of the interaction, Raghavan highlighted his experience in scaling technology to benefit humanity. He also mentioned that the startup is already in talks with several companies interested in utilising Sarvam agents. At the event, the startup revealed that their agent is already being integrated into the Sri Mandir app. (Vivek Raghavan & Pratyush Kumar, co-founders at Sarvam AI) Raghavan said there are multiple models that form the backbone of these AI agents. The first is a speech-to-text model called Saaras which translates spoken Indian languages into English with high accuracy, surpassing traditional ASR systems. The second model, called Bulbul, is text-to-speech, offering diverse voices in multiple languages with consistent or varied options depending on preference. The third is a parsing model designed for high-quality document extraction. This model addresses common issues with complex data, aiming to improve accuracy in parsing financial statements and other intricate documents. Notably, these models are closed-source and available to customers as AI. However, the startup also launched an open-source, two billion-parameter foundational model trained on four trillion tokens and completely from scratch. At the event, the startup also demoed what their agents could do. The demo, which was pre-recorded, showcased how a Sarvam agent could comprehend a person's health condition, assist in finding the right doctor, and even book an appointment. A pre-recorded demo may not appeal to everyone, but from the startup's perspective, it's a safe bet and completely understandable. Live demos carry inherent risks; for instance, at the Made by Google event, one Googler's attempt to showcase Google Gemini's capabilities live saw them fail twice before succeeding. Sarvam AI's demo was also reminiscent of OpenAI's showcase of their latest model, GPT-4o, earlier this year. While Sarvam AI's demo was less dramatic and also not at all controversial, it effectively demonstrated that their agents could understand the context as well as various Indian languages and dialects. "These agents can also be very contextual. For example, when you're on a particular page, you press a button seeking more information about a particular item. The agent will be context-aware, so it knows where you're asking from. In contrast, when you call a number, it starts from scratch without that context," Raghavan said. The startup revealed it trained its models using NVIDIA DGX, leveraging Yotta's infrastructure. Other notable collaborators include Exotel, Bhashini, AI4Bharat, EkStep Foundation and People+ai.

Sarvam launches GenAI platform for India

Bengaluru: Sarvam AI, a Bengaluru-headquartered startup founded last year to develop Indian language generative AI platforms and applications, on Tuesday announced a slew of business-to-business products built on top of its GenAI models. The launch saw the presence of a host of industry bigwigs and government officials, underscoring the importance of the project to make AI accessible to the masses. Infosys co-founder and founding chairman of UIDAI (Aadhaar) Nandan Nilekani was there, as was Nvidia's managing director of South Asia Vishal Dhupar. Many others sent virtual messages. These included Khosla Ventures founder Vinod Khosla, who is an investor in Sarvam, Aadhaar's former chief architect Pramod Varma, Meta vice president of public policy Shivnath Thukral, and Yotta Infrastructure Solutions founder and CEO Sunil Gupta. Nilekani noted that India is a country where AI is being used a lot. "Sarvam AI has now shown that we have the tools and infrastructure to do it at scale and make it cheap. We need to have frugal innovation in AI in India," he said. This is the first time that Sarvam, previously working out of IIT Madras, has unveiled a set of subscription software products for enterprises for commercialisation. Its products also included a mix of open source software. Previously, in December 2023, the company had released its first open-source Hindi language model called OpenHathi-Hi-0.1. It was the first AI model aimed at innovating in AI for Indian languages. Catering to industries including financial services, legal services, consumer goods, technology, media and telecom, Sarvam's product mix unveiled on Tuesday included Sarvam Agents, Sarvam 2B, Shuka 1.0, A1 and several Indic languages-based Sarvam models. Sarvam Agents are voice-enabled, multilingual, action-oriented, custom business agents deployable via telephone, WhatsApp, or in-app. These are available in 10 Indian languages, available at a cost starting Re 1 per minute. 2B is a foundational Indic large language model. Shuka 1.0 is an open-source audio language model, an audio extension on the Llama 8B model to support Indian language voice in and text out, which is more accurate than frontier models. A1 is a GenAI software designed for lawyers to enhance capabilities with features such as regulatory chat, document drafting, redaction and data extraction. Among the Sarvam models, which are available as application programming interfaces (API), were the likes of Maurya for translation, Bulbul for text to speech, Saaras for speech to text and Sesame for bank and account statements. "Our mission is to democratise AI and make it accessible to every Indian, regardless of their linguistic and socio-economic background. This platform is a testament to our commitment to bridging the digital divide and fostering innovation in India's AI landscape," Pratyush Kumar, cofounder of Sarvam AI, said. Kumar is also a founder of a research initiative on open-source Indian language AI, called AI4Bharat, at the IIT Madras campus, where Sarvam AI's office was previously housed. Kumar and cofounder Vivek Raghavan moved to an office in Bengaluru in 2024.

AI startup Sarvam AI unveils mix of open source, enterprise products

Sarvam AI, an AI startup from Bengaluru, introduced new generative AI products. Key offerings include Sarvam Agents, Sarvam 2B, Shuka 1.0, Sarvam Models, and A1. Sarvam will partner with the likes of Yotta, Nvidia, Microsoft Azure, Google Cloud Platform (GCP) and Exotel for these services and products. Bengaluru-based artificial intelligence startup Sarvam AI on Tuesday launched a slate of products, both for enterprise usage and open source, as part of its full-stack Generative AI (GenAI) platform. The firm, backed by investors such as Lightspeed and Peak XV (formerly Sequoia India), said its products will be able to perform various tasks and be voice-enabled, supporting 10 Indian languages, namely Hindi, Tamil, Telugu, Malayalam, Punjabi, Odia, Gujarati, Marathi, Kannada and Bengali. Sarvam cofounder Vivek Raghavan told ET on Monday that the company had collaborated with leading technology and industry partners to develop and deploy this full-stack GenAI platform. These products includeSarvam Agents, Sarvam 2B, Shuka 1.0, Sarvam Models, and A1. The first product, SarvamAgents, will offer multilingual voice-enabled 'agents' that will be able to talk to customers over phone call, WhatsApp or in-app chat. They will also be able to take actions and make decisions based on the inputs provided by customers. The voice agents will be available, starting at Rs 1 per minute, for enterprises across sectors like financial services, legal services, consumer goods, and technology, media and telecom. The second product, called Sarvam 2B, will be an open source large language model (LLM). Sarvam says the LLM has been trained on an internal dataset of 4 trillion tokens that will be able to perform specific tasks in 10 Indian languages with great efficiency. The third product, Shuka 1.0, will be an audio extension on tech major Meta's open source Llama 8B language model to support Indian language usage. This will also be an open-source product. The Indic language models used to build Sarvam Agents will also be available as a product, called 'Sarvam Models'. These models will now be available as application programming interfaces (APIs). These will include models for translation, speech recognition, speech synthesis, and document parsing, and will be available as part of Sarvam's own API platform for developers. The fifth product, called 'A1', is a generative AI workbench designed for lawyers to enhance their capabilities with features such as regulatory chat, document drafting, redaction and data extraction. Raghavan told ET in an interview on Monday that while Sarvam Agents, Sarvam models, and A1 are revenue generating products, Sarvam 2B and Shuka 1.0 will be open-sourced. Sarvam will partner with the likes of Yotta, Nvidia, Microsoft Azure, Google Cloud Platform (GCP) and Exotel for these services and products. Nvidia's DGX infrastructure will power the Sarvam stack, said Vishal Dhupar, the managing director of South Asia for Nvidia,at the event on Tuesday. While Azure, GCP, and Yotta will act as the company's cloud partners, Exotel will power Sarvam's chatbot for its customer conversations. Raghavan on Tuesday also announced that Sarvam will be partnering with Beckn Foundation and fintech firm Pine Labs. The Beckn Protocol powers e-commerce networks like the government-backed Open Network for Digital Commerce (ONDC). Sarvam's launch comes just months after its Matrix-backed competitor Krutrim AI announced a slew of services, including offering its graphics processing units (GPU)-as-a-service and access to its large language models and other open source models hosted on its cloud. Krutrim, part of the Ola group headed by Bhavish Aggarwal, also launched standalone mobile apps for the Krutrim AI assistant to be used by end consumers, in May.

Sarvam AI Launches Full-Stack GenAI Platform With Five Products

Sarvam AI said Sarvam 2B is India's first foundational, open-source, Indic SLM which can perform tasks like translation and summarisation in vernacular languages Bengaluru-based GenAI startup Sarvam AI has launched its full-stack GenAI platform comprising multiple products -- Sarvam Agents, Sarvam 2B, Shuka 1.0, Sarvam Models, and A1. With its diverse products, Sarvam AI aims to revolutionise AI accessibility and adoption across India's diverse linguistic and socio-economic landscape. Speaking to Inc42, Sarvam AI cofounder Vivek Raghavan said, "Our end goal is to enable a billion Indians to access GenAI." He said that companies in the GenAI ecosystem are largely operating in three different areas - building foundational models, fine-turning models, or working at a prompt level. While most companies select either of these three, Sarvam AI wants to have a presence across all three layers, said Raghavan. Here's a quick view of Sarvam AI's five products and their capabilities: Sarvam Agents: These are voice-enabled, action-oriented, custom business agents. Currently, these are available in 10 languages, including Hindi, Tamil, Telugu, Malayalam, Odia, Gujarati, Kannada, and Bengali. These agents are deployable via telephone, WhatsApp, and in-app. The cost of these voice agents starts at INR 1 per minute. Sarvam AI has already deployed these agents for some companies, while it's in talks for more B2B partnerships. From customer service to outbound sales calls, enterprises can use these agents for multiple use cases. Sarvam 2B: The startup claims that this product is India's first foundational, open-source, Indic small language model (SLM). Trained with 4 Tn tokens, Sarvam 2B can perform tasks like translation and summarisation in vernacular languages. Raghavan said that Sarvam 2B will be better at Indian language tasks than international models like Meta's Llama. Shuka 1.0: Sarvam AI is also an open-source model, which the startup claims is the country's first open-source audioLM. Shuka 1.0 is an audio extension on the Llama 8B model that would support Indian language voice in and text out. The startup claims it is more accurate and less cost-intensive than the frontier models. Sarvam Models: Via this product, the startup is making the Indic models, used in the creation of Sarvam Agents, available for use as APIs. It includes models for translation, speech recognition, speech synthesis, and document parsing. A1: This product is particularly designed for lawyers. It's a GenAI workbench designed to enhance lawyers' capabilities with features such as regulatory chat, document drafting, redaction and data extraction. A1 also has tools for the drafting of contracts and share purchase agreements, among others. Commenting on the product launch, Hemant Mohapatra, partner at Lightspeed, one of the investors in Sarvam AI, said, "In a huge and diverse country like India, multi-lingual AI holds the potential to not only bridge the digital divide, but also unlock transformative use cases for 'Bharat'. We are committed to support the team and applaud them for their mission of driving meaningful impact for billions of Indians." Founded in July 2023 by AI4Bharat creators, Vivek Raghavan and Pratyush Kumar, Sarvam AI has been building India-focussed GenAI foundational models and applications in an attempt to enhance the performance and cost-effectiveness of generative AI applications in India. Sarvam AI's five products are based on this thesis. Raghavan said, "There are four main ideas in our thesis - the first is that GenAI in India needs to be in voice in the language of the people. The second idea is that this technology is not for chatting with people but for getting things done." "The third is while there are the likes of Open AI and Anthropic building frontier models, which are excellent, but when it comes to using these models on a day-to-day basis for domain-specific actions, they are expensive and inaccurate. That's when Sarvam AI is building small language models (SLMs) for specific use cases that are more cost-effective, less latent with better output. The last idea is to enable sovereignty," he added. So far, it was expected that Sarvam AI was building large language models and would compete with Ola's Krutrim AI and international majors like Meta, OpenAI, and Google. However, with its entry into the application layer of GenAI, Sarvam AI will face major competition in the GenAI startup ecosystem in India. Sarvam Agents, Sarvam Models, and AI will be the source of the startup's revenue in the coming years. The startup raised $41 Mn (around INR 342 Cr) in its Series A funding round led by Lightspeed Venture Partners, in participation with Peak XV Partners and Khosla Ventures in December last year. It has been the largest fundraise by an Indian AI startup so far. Sarvam AI declined to comment on its current revenue figures. Speaking on its fundraising plans, Raghavan said that the company is "reasonably frugal" in its approach but GenAI is a cost-intensive business. However, he did not share any immediate fundraising plans. As per Inc42's report, GenAI market in India is set to grow exponentially to surpass $17 Bn mark by 2030 from $1.1 Bn in 2023, growing at a CAGR of 48%.

Sarvam AI Launches India's First Open Source Foundational Model in 10 Indic Languages

Called Sarvam 2B, the model is trained on 4 trillion tokens of an internal dataset. Bengaluru-based AI startup Sarvam AI recently announced the launch of India's first open-source foundational model, built completely from scratch. The startup, which raised $41 million last year from the likes of Lightspeed, Peak XV Partners and Khosla Ventures, believes in the concept of sovereign AI- creating AI models tailored to address the specific needs and unique use cases of their country. The model, called Sarvam 2B, is trained on 4 trillion tokens of data. It can take instructions in 10 Indic languages, including Hindi, Tamil, Telugu, Malayalam, Punjabi, Odia, Gujarati, Marathi, Kannada, and Bengali. According to Vivek Raghavan, Sarvam 2B is among a class of Small Language Models (SLMs) that includes Microsoft's Phi series models, Llama 3 8 billion, and Google's Gemma models. "This is the first open-source foundational model trained on an internal dataset of 4 trillion tokens by an Indian company, with compute in India, with efficient representation for 10 Indian languages," Raghavan told AIM in an interaction prior to the announcement. The model, which will be available on Hugging Face, is well suited for Indic language tasks such as translation, summarisation and understanding colloquial statements. The startup is open-sourcing the model to facilitate further research and development and to support the creation of applications built on it. Previously, Tech Mahindra introduced its Project Indus foundational model, while Krutrim also developed its own foundational model from scratch. However, neither of these models is open-source. The startup, which Raghavan co-founded with Pratyush Kumar, also believes that in India, consumers will use generative AI through voice mode rather than text. At an event held in ITC Gardenia, Bengaluru, on August 13th, the startup announced Shuka 1.0-India's first open-source audio language model. The model is an audio extension of the Llama 8B model to support Indian language voice in and text out, which is more accurate than frontier models. "The audio serves as the input to the LLM, with audio tokens being the key component here. This approach is notably unique. It's somewhat similar to what GPT-4o introduced by OpenAI a couple of months ago," Raghavan said. Previously, the startup has hinted extensively at developing a voice-enabled generative AI model. Startups and businesses aiming to incorporate voice experiences into their services can leverage this tool, particularly for Indian languages. Raghavan also said that its aim is to make the model sound more human-like in the coming months. Another interesting development announced by the startup is Sarvam Agents. Raghavan believes that AI's real use case is not in the form of chatbots but in AI doing things on one's behalf. "Sarvam Agents are going to be voice-based, multilingual agents designed to solve specific business problems. They will be available in three channels- they can be available via telephony, it can be available via WhatsApp, and it can be available inside an app," Raghavan said. These agents are also available in 10 Indian languages, and the cost of these voice agents starts at a minimal cost of just INR 1/min. These AI agents can be deployed by contact centres or by sales teams of different enterprises, etc. While these agents may sound like existing conversational AI products available in the market, Raghavan said their architecture, which uses multiple in-house developed LLMs, makes them fundamentally different. "These agents can also be very contextual. For example, when you're on a particular page, you press a button seeking more information about a particular item. The agent will be context-aware, so it knows where you're asking from. In contrast, when you call a number, it starts from scratch without that context," he said. While both Sarvam 2B and Shuka 1.0 are open-source models, Sarvam.ai is making available a bunch of close-sourced Indic models used in the creation of Sarvam agents ready to be consumed as APIs. "These include five sets of models. I will tell you about the three important ones. Our first model, a speech-to-text model, translates spoken Indian languages into English with high accuracy, surpassing traditional ASR systems. The second model is a text-to-speech model which converts text into speech, offering diverse voices in multiple languages, with consistent or varied options depending on preference," Raghavan said. The third model is a parsing model designed for high-quality document extraction. This model addresses common issues with complex data, aiming to improve accuracy in parsing financial statements and other intricate documents. Other announcements made by the startup include a generative AI workbench designed for law practitioners to enhance their capabilities with features such as regulatory chat, document drafting, redaction and data extraction.

Why this AI startup is betting on voice-enabled bots to scale AI adoption in India | TechCrunch

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.