MirrorMine is an AI company focusing on finding and optimizing interrelationships in complex systems. MirrorMine develops intelligent software tools that deliver whole-system improvements in industrial and commercial facilities in hours as opposed to years.
MirrorMine was co-founded by UBC Prof. Scott Dunbar, his former student Ali Hosseinpour and Syniad, on the idea that synthetic data could allow an AI to outcompete a facility operator when optimizing complex inter-related subsystems. Combining recurrent neural networks and supervised learning models, the system was able to achieve tailored recommendations to all subsystems to reduce overall energy consumption of LEED Gold building by a massive 55% after just 4 hours of learning.
- MirrorMine is differentiated by orchestrating whole-system optimization over individual sub-systems using a proprietary meta model approach
- MirrorMine reduces time and cost associated with system optimization, bottleneck identification
- Orchestrating and mining asset level information to provide system level enhancements in facility management
- Green Buildings
- Manufacturing optimization
- Workflow optimization
- Data quality and cost of pre-processing
- Data duplication across asset classes
- Digitizing manual tasks