About
Building the {verification standard} for AI compute on the grid.
Canus Technologies is a for-profit Australia-based company developing an independent measurement and attestation layer for AI compute infrastructure.
As AI data centres become grid-critical electricity loads, regulators, networks, lenders, insurers and infrastructure owners will need better evidence of how these assets actually behave. CANUS is being built to address that evidence gap.
Canus does not provide management software, dashboards, optimization, control, or advisory services to data-centre operators. We do not access workload, customer, model, or interior operational data. Our work begins and ends at the measurement boundary.
Closed Working Group
Canus operates through a closed working group spanning research collaborators at leading universities, senior practitioners drawn from established energy and infrastructure operators, technical advisors, and external patent and corporate counsel, all engaged under non-disclosure as the programme advances toward patent filing and first pilot.
Our public footprint is scoped to to the regulatory record and our engineering work held inside the programme, until patent protection and pilot deployment are in place.
Public record
In May 2026, CANUS lodged a formal submission into the AEMC’s ERC0394 rule-change process on improving National Electricity Market access standards, contributing principles for how AI compute connections should be measured, scoped and verified.
Founder-level engagement with grid operators, infrastructure capital, insurers and AI compute stakeholders is conducted privately under non-disclosure.
Independence by construction
The verification layer for AI compute will be trusted only if the entity producing it has no operational stake in the compute, the energy, or the financial settlement it measures. Canus is constructed to that standard.
Our governance, disclosures, and engagement model are designed to keep that independence intact through commercial scale, which is why each is held to institutional discipline from the first day.
Programme to date
Canus has been operating as a closed working group since 2025. The programme has produced the engineering and policy posture that now anchors our institutional engagement.
Principal Founder
Canus was founded by Siddharth Bedre, whose career spans the power chain that AI compute increasingly depends on.
He has worked across renewable generation, inverter technology, mission-critical power systems and data centre infrastructure, including roles connected to solar inverter markets, UPS-backed power architecture, and large-load energy procurement.
This background gives Canus a practical view across both sides of the problem: the energy systems that power AI infrastructure, and the grid-facing behaviour that third parties increasingly need to trust.
Engagement with Canus
Canus accepts a defined number of conversations each quarter with grid operators, regulators, lenders, and infrastructure investors whose mandates intersect the AI–grid evidence layer.
Inbound enquiries should be directed to research@canus.io with a clear statement of institutional context. We respond selectively.