cbybs-zc1.campaign-view.in Open in urlscan Pro
169.148.148.92  Public Scan

Form analysis
0 forms found in the DOM

Text Content

Join List
Share

0

0

0

0
 
0
0 /250
Post



Express your Like on Facebook

Share this on Facebook

Tell this to your Friends

Say this on X
Tweet
Share this on LinkedIn

Pin this on Pinterest

Share this on Google Plus


Post your comments

0 /250
Post



View on the website »

A few months ago, when we asked Sarvam AI chief Vivek Raghavan if it was too
soon to call the company the OpenAI of India, he humbly declined the comparison.
However, with the recent groundbreaking announcements, it’s clear that the
company has now truly earned the title of ‘Sarvam AI of India.’ Your AI Human,
Amit Raja Naik, is super excited to share this exciting news with you.





India gets its own frontier model, Sarvam 2B 

Sarvam AI’s latest open source model, 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.

With this, Sarvam 2B competes with the likes of small language models (SLMs)
such as Microsoft’s Phi, Meta’s Llama 3 8 billion, and Google’s Gemma among
others. 

“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 ahead of the
release. 

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
(currently available on Hugging Face), while Krutrim also claiming to have
developed its own foundational model (which is currently available on Databricks
Marketplace). 





India gets an AI voice, Shuka 1.0 

Besides releasing open-source LLM, Sarvam AI also released Shuka v1—India’s
first open-source audio language model.

Powered by Saaras v1 (a SOTA, in-house, audio encoder) and Llama3-8B-Instruct,
Shuka v1 is a frugally trained language model that natively understands audio in
Indic languages, excelling in zero-shot QA across multiple Indian languages with
minimal finetuning.

The startup, which Raghavan co-founded with Pratyush Kumar, believes that in
India, consumers will use generative AI through voice mode rather than text. 

“The audio serves as the input to the LLM, with audio tokens being the key
component here. This approach is notably unique. However, it is somewhat similar
to what GPT-4o introduced by OpenAI a couple of months ago,” said Raghavan.

Sarvam AI said that the new model is 6x faster than Whisper + Llama 3. At the
same time, its accuracy across the 10 languages is higher compared to Whisper+
Llama 3.

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. In the coming months, the company aims to make the model sound more
human-like.

(Check out the full interview, exclusively on AIM)





AI agents to the rescue

Another interesting development announced by the startup was 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,” said Raghavan.

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.

Enjoy the full story here. 





(Join Sarvam AI and other industry leaders at India’s biggest generative AI
summit, Cypher 2024, from September 25 to September 27 in Namma Bengaluru.
Register now!)



India’s Fabless Semiconductor Supply Chain is Far from Fabulous



The semiconductor industry in India is fabless, not fabulous, meaning that while
designs are conceived within the country, the actual manufacturing of chips
occurs overseas. While PSMC and Tata Electronic have their roadmap for building
a fab in India by 2026, till then, Indian companies rely on TSCM and others for
building semiconductor chips that were designed in India. Which creates a
security issue within the supply chain of the semiconductor industry.

To explain the possible risks in the semiconductor supply chain and how to
protect from them, Shashwath TR, the CEO and founder of Mindgrove Technologies
spoke with AIM. Check out the full story here. 



AI Bytes

India's answer to Harvey AI, Mumbai-based LexLegis has launched LexLegis AI,
trained on one crore legal documents to streamline research and simplify legal
complexities for professionals.  CoRover.ai, the creator of BharatGPT, has
joined NVIDIA Inception to accelerate the development of India’s first
indigenous generative AI platform, leveraging NVIDIA’s resources to enhance
AI-driven customer engagement across various sectors. Japanese AI startup Sakana
AI has launched ‘The AI Scientist,’ the first system for fully automatic
scientific discovery.  London-based AI startup Cosine AI recently unveiled
Genie, the world’s most capable AI software engineer, surpassing Cognition’s
Devin by achieving a record-breaking 30.08% on SWE-Bench.  Mufeed VH, creator
of the AI software engineer Devika, has launched Asterisk, a YC-backed AI agent
startup revolutionising cybersecurity by autonomously detecting and patching
security vulnerabilities with zero false positives.

AI Conclave Wonders



Cypher 2024 marks a significant expansion as it celebrates its 8th edition by
branching out to the USA in addition to its already established presence in
India.

Browse through the links below to learn more about the different editions of
Cypher 2024.

 * Click here to know more about Cypher India 2024
 * Click here to know more about Cypher USA 2024


These links will guide you to comprehensive event information, including
agendas, speakers, registration details, and more.

Enjoying Sector 6 (formerly AIM Daily XO)? Share it with colleagues or friends –
they can sign up here. 

We love hearing from our readers! Have thoughts on our new format? Questions,
comments, or ideas are always welcome. If there’s a specific topic in AI or
analytics that you're curious about, tell us!

Reach out to us at info@analyticsindiamag.com.
Stay tuned for more insights in our next edition!


Curated with ♥️ in Namma Bengaluru



                                                           
This email was sent by info@aimmediahouse.com to aa.apte@godrejinds.com
Not interested? Unsubscribe | Manage Preference | Update profile

Analytics India Magazine | 280, 2nd floor, 5th Main, 15 A cross, Sector 6, HSR
layout Bengaluru, Karnataka 560102