 "Bhavish Aggarwal")
Krutrim, the artificial intelligence (AI) venture of Ola Co-founder Bhavish Aggarwal, unveiled its vision for the developer community to foster innovation in the Indian AI landscape. Krutrim also announced mobile apps for the AI assistant.
The company, focused on developing full-stack AI capabilities, has opened up its cloud platform to enterprises, researchers, and developers, to enable them to develop new-age technology products at a fraction of current costs.
The cloud platform provides access to AI computing infrastructure, foundational models and open-source models being hosted on Cloud. The platform also provides foundational location services APIs and SDKs to build innovative localised mapping based solutions.
Additionally, the Krutrim assistant app, built on the company’s own LLM (large language model—trained on over two trillion tokens with the largest representation of Indic data), will simplify leveraging the power of AI for everyone. It currently understands and generates intelligent responses in over 10 India languages, which will be expanded to 22 official languages in the future. The app will allow users to give voice prompts and it will integrate text, voice, and visual data in the output. The Krutrim assistant also works with other apps to perform tasks such as booking cabs, setting reminders, or sending messages without having to switch between applications. It also adapts to user’s preferences and habits over time, offering a personalised experience with tailored suggestions and reminders.
Bhavish Aggarwal said the firm is committed to developing full-stack AI capabilities in India, for the world. “We truly believe that India needs its own technology platforms to enable the emergence of world-class products at a fraction of current costs,” said Aggarwal. “With Krutrim, our enterprises and developers will no longer have to be dependent on complex and cost-heavy western cloud platforms, which are not optimised for the Indian use cases. Our Krutrim assistant app will revolutionise the adoption of GenAI with its ability for a seamless integration into everyone’s life by solving the complexity and fragmentation often experienced with digital tools and services.”
With access to Krutrim Cloud, developers will now be able to build scalable products from scratch with the following models, tools, and services:
More From This Section
Model-as-a-Service
Krutrim has announced Model-as-a-Service (MaaS) offering developers access to its LLMs as well as open-source models being hosted on its cloud. The platform’s base LLMs and multiple associated instruction-tuned versions, together with the Python SDKs, will enable developers to create India-specific products, given its superior understanding of Indian languages and cultural context. Moreover, developers and enterprises can benefit from cost advantages owing to much cheaper inference costs, thanks to the platform’s vertical integration of the whole AI solution stack.
The company is also planning to release models for voice, image understanding and generation, and pre-tuned LLM agents.
GPU-as-a-Service
Krutrim Cloud will be offering GPU-as-a-Service on its state-of-the-art AI computing infrastructure, allowing enterprises and developers to train, fine-tune and serve their models.
Foundational services
In an effort to help enterprises focusing on localised location services, the Krutrim cloud is offering the following location API services.
- Places API (Autocomplete and Reverse Geocoding)
The Places API helps consumers find specific locations like restaurants, hotels, or charging stations on a map. It helps get detailed information about these places, such as their names, addresses, and categories.
- Tiles API
The Tiles API helps load and display small sections of a larger map, called tiles, which together form the complete map. This approach makes loading maps faster and more efficient as only the parts of the map one needs to see are loaded.
- Routing APIs
The Routing and Directions API will help generate accurate and optimised routes between multiple waypoints, including detailed turn-by-turn directions, estimated travel times, and alternative routes. This powerful tool empowers users to plan their journeys effectively, whether they are navigating through busy city streets or exploring remote landscapes.
- Navigation SDKs for iOS and Android
Navigation SDKs include tools and code libraries that developers use to build navigation features such as turn-by-turn directions, real-time traffic updates, and route planning, among others, into their own apps.
The company said it envisions creating a comprehensive, reliable, and scalable ‘maps platform’ that addresses local needs and supports broader initiatives for technological and economic advancement within India. The company is also working on developing MapGPT to enable natural conversations powered by location-aware AI intelligent assistants. Further, it aims to offer an immersive experience of the location hotspots besides integrating community-based features for real-time updates on traffic and road conditions and add more features to improve personalisation.
