Dhananjay M.

WORK HISTORY

EDUCATION

ABOUT DHANANJAY M.

I’ve always believed that if AI governance feels like a \"stop sign,\" it’s been designed wrong. The best governance acts like the high-performance brakes on a Formula 1 car—it’s the only reason the driver feels safe enough to go 200 mph.My journey into the \"why\" and \"how\" of responsible AI started at Mastercard. There, I didn\'t just build models; I built the factory. I led the development of the Data Science Workbench, an enterprise AI/ML platform. The challenge wasn\'t just the code—it was building the governance architecture around the platform, ensuring that thousands of data scientists could innovate at scale without compromising privacy or security.Today, I focus on moving governance away from static checklists and into Operationalized Reality. I specialize in establishing multi-tier governance structures—from the boardroom to the development team—to ensure AI strategy is both board-ready and technically sound.I look at risk through the entire AI Lifecycle, mapping every technical vulnerability to global standards like NIST AI RMF and PIPEDA- Data Ingestion: Identifying privacy \"poisoning\" and lineage risks- Training & Tuning: Hard-coding checks for bias and IP leakage- Inference & Deployment: Monitoring for hallucinations and model drift.As we move into the era of Agentic AI, the stakes are higher. Since autonomous systems can be unpredictable, I developed an AI Risk Assessment Tool to automate the complex intake and risk-mapping process. To ensure these systems remain auditable, I advocate for deep observability within the architecture—integrating span and trace capabilities into Agentic and RAG-based systems. This ensures that every autonomous \"thought\" or retrieval is traceable, transparent, and safe.With an M.S. in Computational Mathematics and an MBA, I’m the bridge between the math and the mission. I can dive deep into the nuances of a neural network with engineering teams and then pivot to discuss ROI and risk posture with the C-suite.What I’m building for 2026- Governance-as-Code: Moving compliance requirements directly into the product development lifecycle- Third-Party AI Risk: Vetting the AI supply chain to keep proprietary data and IP secure- Traceable Autonomy: Scaling the observability frameworks needed to govern the next generation of Agentic AI.