Karnataka launches KEO to take AI-ready computing to homes, schools and rural communities: DK Shivakumar

Bengaluru: Karnataka is taking advanced computing beyond technology hubs and into the hands of students, entrepreneurs, farmers, small businesses and households with the launch of KEO, an AI-ready computing platform developed in the state, chief minister DK Shivakumar said on Friday.

“KEO is a Made-in-Karnataka platform that brings AI-ready computing closer to students, entrepreneurs, farmers, small businesses and institutions, including those in areas where connectivity may be limited,” Shivakumar said in a statement.

The platform, he said, would help move the state from digital access to “meaningful participation in the digital economy” by making advanced technology homegrown, open and accessible.


KEO, which stands for Knowledge-driven, Economical and Open-source, combines an open-source RISC-V processor architecture, open-source hardware and software and on-device artificial intelligence in a compact desktop platform.

Also read | America sold the world a dream – and kept millions chasing it

Unlike conventional systems that depend heavily on cloud connectivity for AI applications, KEO has a dedicated on-device AI accelerator that allows AI workloads to run locally. This makes the platform suited to schools, institutions and communities where reliable internet connectivity may not always be available.

As part of the launch, 3,000 KEO devices will be rolled out. The first batch of 5,000 devices is already ready, according to the state government. Twenty-three pre-booked customers received their devices from the chief minister at the launch.

The government plans to scale up deployments in the coming months, taking KEO into more schools, institutions, businesses, households and citizen-facing environments across Karnataka.

The KEO portfolio will also be expanded with KEO Lite and KEO Pro. KEO Lite is being developed as a RISC-V-based alternative to the Raspberry Pi 5, with prototypes expected by October. KEO Pro will be a more powerful next-generation version aimed at more demanding computing and AI applications.

Two MoUs were also signed between the Shikshana Foundation and the Karnataka Residential Educational Institutions Society (KREIS), and between the foundation and RSA, to support KEO's deployment across educational and community settings.

Under the Phase 1 rollout supported by the Shikshana Foundation, 2,222 KEO systems will be deployed across all 31 districts and 239 talukas of Karnataka.

Also read | Tata, Godrej and now Birla: Big business groups aren't just chasing glitter

Two Arivu Kendras in each of the 239 talukas will receive four KEO systems each, accounting for 1,912 systems across 478 centres. Another 310 systems will be deployed across 31 residential schools, with each school receiving 10 systems.

The systems are aimed at giving students, educators and communities access to computing, digital learning and AI-enabled educational tools, including in areas with limited connectivity.

IT/BT minister Priyank Kharge said KEO was the state's “practical answer to the digital divide”, adding that affordable mass computing would enable students, small businesses and households to participate more fully in the digital economy.

KEO comes preloaded with BUDDH, a homegrown AI learning assistant trained on the Karnataka State Education Research and Training (DSERT) syllabus. The assistant provides interactive, syllabus-aligned learning support and can operate without an internet connection.

The platform has been developed through a collaboration involving the Department of Electronics, IT and Biotechnology, Government of Karnataka, KEONICS, semiconductor partners and AI-focused system integrators.

KEONICS chairman Sharath Kumar Bache Gowda said the platform was intended to create an adaptable computing ecosystem that universities, startups and businesses could build upon.

“KEO is not just a device; it is the beginning of a larger computing ecosystem,” Gowda said, adding that the first few thousand units were only the beginning of the deployment.