An open AI research initiative: we study how autonomous AI agents coordinate in industrial environments — and publish the standards, code and findings.
Negotiation, task allocation and conflict resolution in decentralized agent contours under real-time constraints.
Formal contracts, auditability and interoperability requirements for multi-agent systems in regulated industries.
Methods for rebuilding legacy industrial software — telematics, dispatch, compliance — as agent-based decision systems.
Monitoring and decision-support loops where human operators and autonomous agents share responsibility.
An open framework for building auditable industrial agent systems: agent contracts, communication protocol, orchestration semantics. Free to use, fork and implement.
Every CIRIYC PRO contour is built on MAIF. The framework is registered, versioned and published in full — specification, reference architecture and examples.
github.com/ciriycpro/MAIFFirst-edition ПНСТ draft: terminology, architecture requirements and conformance criteria for industrial MAS. Developed within TC 164.
Case study of re-architecting legacy field-workforce telematics into a real-time multi-agent decision system. Industrial validation in progress.
C4 models, architecture decision records and service blueprints for reproducible multi-agent system design.
React-based interactive prototype of the field-workforce agent contour.
Multi-agent conversational layer for industrial lead qualification, running on ciriyc.ru and ciriycpro.com.
We study multi-agent coordination as a practical path toward general artificial intelligence in the physical economy. Superintelligence, in our view, arrives grounded: auditable agents, open standards, verifiable behaviour — intelligence you can certify, not a black box.
