In partnership with

Enterprise control layer for registering, monitoring, and governing AI agents

Aigentsphere was founded in Sydney in 2025 by Kelly Bayer Rosmarin, formerly chief executive of Optus, and Quinton Anderson, previously an executive at Commonwealth Bank and Optus. The company frames agent sprawl as a workforce management problem rather than purely a security one: agents are deployed independently by engineering, finance, and marketing teams, they consume budget, they produce work of variable quality, and no single function can say how many exist or what they are costing. The platform's answer is a registry and an accountability layer over the top, applying the practices organizations already use for people to a population of software workers.

Agents are onboarded through a REST API, OpenTelemetry instrumentation, or an embedded LLM gateway, with support for LangChain, OpenAI, Anthropic, and AWS Bedrock. Once connected, the platform runs configurable risk indicators with tiered effectiveness ratings and automatic alerting, tracks cost and operational KPIs on dashboards aimed separately at executives, compliance officers, and the engineers who build the agents, and applies LLM-as-judge assessors that score outputs against defined quality standards without a human reviewing every response. A policy library with templates for GDPR, the EU AI Act, and ISO standards drives automated assessment and generates the audit trail. Aigentsphere raised a seed round of about A$4 million, reported as roughly US$2.6 million, led by Main Sequence, the deep tech fund co-founded by CSIRO.

Market Segments:

AI SecurityGovernanceCompliance

Categories:

AI GovernanceAI AgentsCompliance AutomationRisk Management