- Top financial institution was faced with increasing costs and resource needs to comply with increasing regulatory requirements
- Complex nature of organization creates challenges for transparent monitoring and enforcement
- Steady increase of human capital-intensive compliance costs
- Environment of record penalties and market cap destructive and career-ending reputation impact
-The world’s top 20 banks have paid more than $235bn in fines since the financial crisis including $141bn for US mortgages, $44bn for UK customers, $14bn for business with sanctioned countries
-Fines and penalties equivalent to the annual economy of Greece or Portugal
- Regulatory and compliance application developed with Lucid.AI
- Comprehensive platform identified employees possibly trading on insider information
- Lucid.AI was able to see connections across long chains of relationships not detected by human compliance functions
- Ability to synthesize across multiple data sources, finding long connection chains which properly weight the perceived significance of links including:
- Recommendations and conclusions can be reviewed by stepping through the full logical sequence; experts extend the app by editing knowledge, not code
- Eliminates many of the false positives and false negatives which were generated by their previous so-called “deep learning” algorithmic solution
- Preemptive measures created meaningful value and avoided potential fines, and avoided diminishing brand equity and market value
- Commitment to invest the R&D necessary to develop additional applications on the Lucid.AI platform
- Discussions regarding a strategic partnership to support the Lucid ecosystem, including a multi-year revenue commitment to support continued Lucid.AI R&D
- Multinational oil company is faced with high costs of oil and gas extraction, both from rich deep water deposits and from poor “oil sands” deposits
- The need to extend conventional oil and gas reserves have resulted in more widespread deployment of technologies such as ultra-deep-water drilling, hydraulic fracturing and land-based gravity-assisted steam generation
- Developed an overlay to well monitoring software to enhance system engineers’ ability to drive well head efficiency
- The application was built not by programming, but by incrementally teaching Lucid.AI about the engineering and science involved, conditions of interest, etc., and giving it detailed models and histories of each individual well being monitored
- Lucid.AI applied its new domain expertise to the incoming streams of real-time well data, formulating plausible hypotheses about what geophysical and geochemical conditions of interest might be occurring down-hole, enabling much more rapid response to acute problems, anomalous data, and in some cases identifying chronic problems months earlier than the previous solution (verified by running on historic data)
- Ability to reason about what the data might mean, rather than applying rigid pre-conceived programmed criteria. This often led to alerts which would otherwise have remained masked until some later time, and also led to the suppression of a large fraction of false alarms which would otherwise have wasted staff engineers’ time and led to dangerous “warning fatigue”
- Reduced the need for expensive unnecessary shutdowns (what would have been false positives) and saved equipment from damage and increased total lifetime well yield by earlier detection and identification and analysis of acute and chronic conditions (what would otherwise have been false negatives)
- “Can increase yields 10+% on all deep-water platforms” – J.B., Chief Engineer, Platforms, at that Multinational oil company
- Application allows engineers to view potential issues as they occur, explains its line of reasoning behind each alert (and decision not to alert) and provides recommended solutions to each problem
- Discussions regarding a strategic partnership to support the Lucid ecosystem
Energy Forecasting (Smart Grid)
- Develop a real-time forecasting system for energy distribution networks to make low carbon technologies affordable, sustainable and competitive based on analysis of large quantities of data on network topology and devices, energy demand and consumption, environmental data and energy prices data.
- Manage energy offer / demand, decision making across network monitoring, anomaly detection, route cause analysis, trend detection, planning and optimization.
- Lucid.ai is a complex-event processing platform combining multimodal data integration and fusion, predictive analytics, context- based reasoning, forecasting of consumption, production and price and on-going learned knowledge generation.
- Lucid.ai is able to control, manage, analyze and predict behavior in an extensible manner on electric power grids.
- Generic framework extensible to other networks like gas distribution, steam heat distribution, and water supplies.
- Field trialed in publicly owned buildings in Turin Italy, district heating management in the city of Reggio Emilio, smart electric car chargers in Aachen, on a University campus in Athens and for municipal streetlights in Slovenia.
- Demonstrable energy savings, functional trading system, grid optimization with socially visible economic impact.