Model Training for Regulated AI
Designed for CCOs, CROs, model risk professionals, and data science teams in regulated financial services. You will walk away with a shared framework for evaluating training risk.
What you'll learn
- How training data quality, labeling, and representativeness create downstream compliance risk
- Why feature selection decisions carry proxy variable risk, disparate impact exposure, and documentation expectations
- What examiners expect to see in your training records for reproducibility and validationHow to build training validation workflows aligned to SR 11-7 / OCC 2011-12
- Where training governance connects to your CMS, model inventory, and ongoing monitoring
Why this matters now
Examiners increasingly evaluate how models are built: data sourcing, bias testing before production, feature selection rationale, and reproducibility. At the same time, transfer learning, synthetic data, and automated feature engineering are introducing risk categories most compliance frameworks have not addressed. Organizations that close this gap reduce fair lending exposure, strengthen exam readiness, and build models that are defensible from the start.
Speakers
Amber de Volk
As a senior compliance architect and regulatory advisor Amber has over eighteen years of experience across fintech, banking, payments, and emerging financial technologies. She holds an LLM in Financial Risk Management and specializes in CMS and AML program development, model governance, AI oversight, and enterprise risk management. Known for her strategic approach and hands-on leadership style, Amber partners with founders and executive teams to build scalable, defensible compliance frameworks that drive growth and enable safer innovation.
Uriel Ash
Uriel Ash is the founder and president of Nathan Ash & Associates. As a fractional CIO and CISO with over two decades of experience across fintech, banking, payments, and cybersecurity, Uriel specializes in security architecture, regulatory compliance, and building high-performance engineering teams. A Columbia University-trained electrical engineer and Certified Information Security Manager (CISM), Uriel previously served as SVP and Head of Innovation, IT Product, Security Architecture, AI, and Blockchain at Cross River Bank, where he oversaw more than 20 global development teams and led the creation of systems that powered billions of dollars in lending. Known for his hands-on technical leadership and ability to scale secure, compliant infrastructure, Uriel partners with banks, fintechs, and regulated platforms to build technology strategies that meet the highest standards of security and regulatory readiness.