Ziroh Labs Launches Kompact AI to Power Enterprise AI Workloads on Ubiquitous CPU Infrastructure
Deep-tech and AI startup Ziroh Labs has announced the commercial launch of Kompact AI, a CPU-first AI software platform designed to run enterprise-scale artificial intelligence inference across existing, ubiquitous CPU hardware.
The platform addresses a critical operational reality across global enterprises: while massive GPU clusters remain necessary for training frontier foundation models, the vast majority of operational enterprise AI workloads, including automated software testing, document parsing, customer service routing, cybersecurity telemetry, and domain-specific automation, rely on smaller models and process high volumes of data where I/Othroughput, execution latency, data locality, and predictable infrastructure costs matter far more than raw floating-point compute.
The launch event featured Dr. Whitfield Diffie, pioneering cryptographer and recipient of the 2015 ACM A.M. Turing Award, alongside prominent academic and industry figures including Prof. S. Sadagopan, Founder Director of IIIT-Bangalore.
Decentralising AI: The Turing Award Perspective
Dr. Whitfield Diffie, widely renowned for his foundational contributions to public-key cryptography and decentralized architectures, voiced strong support for Ziroh Labs’ approach. Viewing AI as one of the most transformative technologies since the inception of the World Wide Web, Dr. Diffie emphasised that the societal and economic impact of artificial intelligence will be dictated by how widely its execution layer can be distributed rather than remaining confined to centralised, specialised compute silos.
Commenting on the platform during the launch, Dr. Whitfield Diffie noted:
“The ability to run AI on ubiquitous CPUs is one of the most exciting things I have seen in quite a number of years.”
Purpose-Built Kompact AI Boxes for Regulated Verticals
Rather than functioning solely as an isolated execution runtime, Kompact AI is packaged into sector-specific appliances termed Kompact AI Boxes. These pre-configured software and hardware stacks combine optimised models with localised inference pipelines tailored for distinct industry requirements:
- Fintech & Banking: Secure, low-latency financial transaction analysis, automated fraud pattern scoring, and local credit document validation.
- Healthcare & Clinical Systems: Privacy-preserving electronic health record (EHR) summarisation and medical imaging triage deployed directly on local hospital servers.
- Retail & Supply Chain: Edge-based inventory routing, computer vision for store operations, and conversational customer support.
- Sovereign & Defence Deployments: Air-gapped, fully autonomous AI execution environments ensuring zero telemetry or sensitive corporate data leaves internal enterprise perimeters.
Democratizing AI in Classrooms and the IIT Guwahati Math Lab
To build domestic research capabilities and eliminate resource barriers in academia, Ziroh Labs has embedded a wide-reaching educational charter into Kompact AI:
- Free Academic Licensing & CoEs: Providing academic institutions with free software licenses and setting up Centres of Excellence (CoEs) to grant students direct access to practical AI development environments regardless of institutional funding levels.
- IIT Guwahati Math Lab Initiative: Partnering with researchers and faculty at IIT Guwahati to establish the Math Lab, a fundamental research initiative investigating whether mathematical representations within modern Large Language Models (LLMs) and State-Space Models (SSMs) provide structural foundations capable of advancing toward general machine intelligence.
- 40+ Hour Certified LLM Curriculum: Co-developing a comprehensive 40-hour open course alongside IIT Guwahati covering advanced LLM architectures, accessible for free globally, with optional institutional certification.
Executive & Academic Perspectives
Hrishikesh Dewan, Co-founder and CEO of Ziroh Labs, emphasised accessibility and cost control:
“The next phase of AI adoption will require infrastructure that is accessible, efficient, and closer to the people and businesses using it. Our focus is to make AI more accessible by enabling organisations to run AI where it is needed, with greater control over cost, performance and data.”
Prof. S. Sadagopan, Founder Director of IIIT-Bangalore, praised the deep-tech engineering and frugal innovation:
"Sharing the stage with Turing award winner is such an honour. I congratulate Ziroh Labs for doing so much so soon...10+ partners, 100+ technical papers, 1,000+ models being released is incredible. Kompact AI is creating a huge ecosystem for deep tech. Frugal innovation and building energy-efficient AI for everyone is what excites me about the platform which is a timely development for India."
By turning existing enterprise CPU infrastructure into efficient AI inference engines, Ziroh Labs provides organisations with a cost-effective, privacy-compliant foundation to scale AI across everyday business workflows.
Read More:
2 Min read