ModelOp: Enabling Faster and Powerful Model Governance

Pete Foley

Co-founder & CEO

“ModelOp Center automates the governance, management and monitoring of deployed AI, ML models across platforms and teams, resulting in reliable, compliant and scalable AI initiatives”

As the AI market was evolving, two entrepreneurs, Pete Foley and Stu Bailey talked with many large, non-digital native enterprises and discovered there was a gap that wasn’t being addressed. That gap was the need for large enterprises to operationalize diverse types of models — not just new, AI models. In 2016, ModelOp was founded to address this gap, creating a team with strong competencies in data science, software engineering, business process, risk management and compliance. ModelOp Center is the result of the early ModelOp team working with some of the largest global banks and financial institutions and helping them accelerate the operationalization of their models — all types of models — and ensure that governance and regulatory requirements were strictly enforced and auditable.

Today ModelOp Center is recognized by analysts as a leading ModelOps software solution and core to any AI orchestration platform. Large organizations use ModelOp Center to govern, monitor and manage models across the enterprise and unlock the value from their analytic, machine learning and AI investments. ModelOp Center automates the governance, management and monitoring of deployed AI, ML models across platforms and teams, resulting in reliable, compliant and scalable AI initiatives. Organizations that use ModelOp Center accelerate model operationalization on average by 50% and reduce operationalization costs by 30%.

AI model governance starts at the time of ideation and use case definition and ends at model retirement. Successful AI model governance requires a set of standards and processes that are adhered to throughout the model life cycle – development, validation, productionization, and operations to retirement. The company’s set of predefined automated processes make it fast and easy to define and establish business, risk, and compliance rules for the entire model life cycle. A single model inventory for all AI and analytic models, regardless of type or where they are run, give you the control to apply standards and rules across all models in the enterprise. Moreover, these models are continuously learning, which often leads to changes in model decisioning. Operational models must be constantly checked for accuracy. If changes in decisioning or operational performance are not detected and immediately resolved, unreliable decisioning can put the business at risk. Equally important is understanding what changed and why.

A successful AI strategy requires the ability to quantify the ROI of AI projects and the models that drive them. Without the proper insights, determining if models are meeting their intended business objectives and measuring ongoing business value, is difficult to impossible. ModelOp Center AI Operational Assurance automates end-to-end management and governance of all AI and decision-making models across the enterprise. The solution’s comprehensive and persistent data retained on each model throughout its life delivers the reproducibility that is often critical for passing audits and adhering to AI regulatory guidelines. Integrations with Tableau and Power BI give the flexibility of custom reporting so users are ready and able to pass any audit at any time.

An interesting insight that highlights the company’s value proposition is when the team assited the Royal Bank of Canada. Bond pricing is a real, long-standing challenge for traders. Making profitable trades hinges on being able to accurately calculate bond price movements, which can be achieved by utilizing AI models as part of the decisioning process. However, for traders to adopt AI models, it was imperative that all models were explainable at each moment to traders and business executives. Royal Bank of Canada implemented a comprehensive ModelOps capability to support its municipal bond trading operations by providing transparent, interpretable, and easily accessible information that allows traders to make confident recommendations. With this successful pilot, the team embarked on an AI operationalization strategy to provide consistent AI model deployment, scoring, governance, and monitoring across the enterprise.

Today, ModelOp helps businesses accelerate time from model deployment to decision making by 50% or more by automating the entire operationalization process and with out-of-the-box integrations. It also helps increase model contribution by 10% or more using automated, reflexive monitoring that maintains accurate model decisioning while enabling to increase the IT and model operations team productivity by 50% or more with pre-defined, automated model life cycles and out-of-the-box integrations.

With over 25 years of executive and entrepreneurial experience in enterprise software and a track record of successful business exits, Pete’s leadership gives ModelOp customers, partners and employees a high level of trust and confidence in the company and its future.