Scotiabank - Global Banking and Markets

Recently, I was proud to welcome our leadership team to Texas for our first-ever GBM Senior Leadership Forum. We hosted discussions on how we will continue to raise the bar for our clients, improve global connectivity, and drive high performance through how we lead and show up. Thank you @Chad Wallace for joining me on stage for a thoughtful conversation on ways we can continue to connect Global Transaction Banking with Corporate Banking to better serve our clients.

The most valuable lessons from my co-op had nothing to do with code. After a year on Scotiabank 's Algorithmic Trading team, here are 4 lessons I wish someone had told me on day one: 1. Network inside your company. Not the kind that gets you a job. The kind where you go into conversations with a genuine interest in learning what others do. You have proximity to incredible teams and people, so take advantage of that opportunity. I found mentors, understood the ecosystem that my work fit within, and learned about the depth of opportunities in my company. It later proved invaluable when I assisted another team on a system similar to one I'd built. 2. Learn from people outside your expertise. Some of my best growth came from conversations with traders and salespeople on the trade floor. Not about code, but about how to project confidence and speak with conviction. Turns out, these skills translate to demos and stakeholder updates just as well as client calls. Stepping outside your bubble lets you pick up insights that you can't find in your own field. 3. Document everything from day one. Write down how you got set up, what you're learning, and the processes you follow. Writing it down forces comprehension, not just absorption. And if no setup documentation exists, congratulations! You've just created the first version of a guide that will help every future hire. People are even building AI tools to automate documentation. That's how universally necessary it is. 4. Assume that you're wrong. Prove to yourself that you're right. Before touching anything critical, I wrote down every step and planned for what could go wrong. During an upgrade to the data application powering our trading algorithms, that habit helped me catch issues early, stay confident, and earn trust from my team. The method matters less than the mindset. Assume you're wrong until you've proven otherwise. The technical skills were important, but these habits are what I'll carry with me into every future role. If you're starting a work term, I hope they help you too. What's a non-technical lesson that stuck with you? I'd love to learn from your experience :)

Participating in the CARIFIN “All Urban” Challenges 2026 has been not only a rewarding one but reflective experience. This year, I set a personal goal—not only to complete the Green Mile and the Chancellor Hill Challenge, but to improve on my time from 2025 and push myself a little further. While personal progress remains important, what stood out most was the opportunity to share the experience with an incredible team. Goal achieved and exceeded 😁💪👌 I also had on opportunity to invite a few guests, my Regional Director, Jesus Emmanuel Reyes, Senior Manager Leonel Espinaz Perez, and Senior Manager Intake North & Central, Claudia Sanderson, to join me for the Chancellor Hill Challenge. Their participation added a special dimension—bringing to life true team spirit, collaboration, enthusiasm, passion, and drive. It was memorable to see leaders come together outside of the workplace to embrace the challenge, support one another, and lead by example. These moments reinforce that success is not just about individual milestones but about the strength we build collectively. I’m pleased they could have joined and be part of this last activity. I reflect on the energy, camaraderie, and determination that made this challenge so special. Stronger together—on every hill we climb! #CARIFIN #Leadership #Teamwork #Collaboration #PersonalGrowth #Wellbeing #StrongerTogether #EmployeeEngagement #ChallengeYourself #Inspiration

Released April 17, 2026: The Federal Reserve Board has redefined the architectural baseline for Model Risk Management. With SR Letter 26-2, the Fed, Federal Deposit Insurance Corporation (FDIC) , and Office of the Comptroller of the Currency shift toward dynamic, risk-based oversight for organizations with >$30B in assets. This marks the end of "check-the-box" validation. The Ten Pillars of SR Letter 26-2 (What Matters and What You Should Know) 1. Effective Challenge: Requires objective experts with authority to halt deployment. 2. Risk-Based Tailoring: Supervision scales with model complexity, not just presence. 3. Inherent Risk: Focus on theoretical constraints of assumptions and data. 4. Materiality: Aligning model exposure directly with institutional purpose. Conceptual Soundness: Mandatory "first-principles" assessment of design. 5. Outcomes Analysis: Rigorous back-testing against real-world volatility. 6. Continuous Monitoring: Shift from annual reviews to real-time relevance. 7. Aggregate Risk: Managing hidden dependencies across the inventory. Vendor Accountability: Principles apply even to proprietary "black boxes." 8. Documentation: Ensuring oversight survives personnel turnover. The Strategic Red Lines 9. Exclusions: Does not apply to banks <$30B or rule-based spreadsheets. 10. The AI Gap: Generative and Agentic AI are currently out of scope, requiring separate National Institute of Standards and Technology (NIST) AI RMF governance. This creates a dangerous "governance bifurcation" for the C-Suite. Institutional Readiness ✔️Inventory: Transform static lists into dynamic risk-intelligence repositories. ✔️The Talent War: Secure teams with the technical depth and independence to challenge the business line. ✔️Third-Party Calibration: Update SLAs to ensure access to critical validation data. Three Questions for the Board ⟶ Does "Effective Challenge" have the authority to stop a model, or is it a performance? ⟶ How are we managing the aggregate risk of shared data dependencies? ⟶ Is our materiality assessment aligned with the 2026 volatility landscape? The Regulatory Intersection SR 26-2 should be harmonized with the National Institute of Standards and Technology (NIST) AI RMF, DORA, and ISO - International Organization for Standardization /IEC 42001. Compliance fragmentation is the primary driver of modern operational failure. Success now depends on the ability to harmonize disparate standards into a single, defensible source of truth. #ModelRisk #FedReserve #GRC #RegulatoryStrategy #RiskManagement

📉The Illusion of Control in 2026 AI Governance The Reality Check: The latest joint assessment from the Bank of England and Financial Conduct Authority reveals a dangerous divergence in UK financial services. While 75% of firms have scaled AI adoption, nearly half (46%) admit to only a "partial understanding" of the very models they deploy. The industry is currently hiding behind the "Human-in-the-Loop" (HITL) safety net—with only 2% of use cases being fully autonomous. But as agentic AI accelerates, "HITL" is becoming a bottleneck, not a guardrail. The "Sovereign" Analysis: 3 Warnings for the C-Suite Based on the 2026 supervisory priorities, your governance must pivot from "Wait-and-See" to "Evidence-at-Scale": ⟶ The Accountability Trap: 84% of firms have a "Named Accountable Person," but accountability without technical observability is just a liability. If you can’t audit the "why" behind an AI decision, your Senior Manager status is at risk. ⟶ The Concentration Crisis: 33% of AI use cases are now dependent on a handful of third-party providers. We are creating a systemic "Single Point of Failure" that the Critical Third Party (CTP) Regime will target by year-end. ⟶ The Complexity Debt: "Model Complexity" is now cited as a Top 3 risk. We are layering advanced LLMs on top of legacy data foundations. In 2026, complexity is the primary vector for model drift and adversarial injection. The 2026 Execution Framework: Moving to Agentic Governance To clear the FCA’s AI Live Testing hurdles and align with NIST AI RMF, firms must transition to: ✔Real-time Explainability: Moving beyond "Post-hoc" reporting to "Runtime" transparency. ✔Agentic Action Boundaries: Defining where an AI's authority ends and the human mandate begins. ✔Interoperable GRC: Anchoring your AI risk directly into National Institute of Standards and Technology (NIST) CSF 2.0 and ISO 42001. The Bottom Line: High-end AI adoption is being chilled by "accountability confusion." The winners of 2026 won't be the ones with the fastest models—they will be the ones with the most auditable guardrails. #AIGovernance #FinancialServices #RiskManagement #BankOfEngland #FCA2026

🏛️ The 2026 Gap: Capability vs. Readiness | Stanford University HAI AI Index Report A synthesis of this 400-page document reveals a critical divergence: AI acceleration is fundamentally outstripping institutional capacity. We have entered a period of Systemic Stress Testing. 🚀 Top 5 Strategic Shifts 1. Model Convergence: U.S. and Chinese performance has converged. Competitive advantage now resides in Execution Excellence. 2. Labor Inversion: Productivity has increased 14-26%, yet entry-level developer roles declined 20%. This poses a risk to the 2030 talent pipeline. 3. Compute Sovereignty: Given supply chain fragility, AI infrastructure is now a matter of national security. 4. The "Jagged Frontier": Models solve complex physics but fail at basic reasoning, such as reading analog clocks (50% accuracy). This is a primary fiduciary risk. 5. Industrial Footprint: A single training run emits 72k tons of CO2. AI is now a heavy-industrial utility. ⚖️ The Fiduciary Crosswalk 1. Policy: Organizations must navigate a "Dual-Speed" environment as the EU enforces regulations while the U.S. deregulates. 2. Safety vs. Accuracy: AI incidents have increased 362%. Implementing safety filters may inadvertently diminish model accuracy. 3. Quantum Privacy: Without National Institute of Standards and Technology (NIST) PQC standards, data is vulnerable to "Harvest Now, Decrypt Later" strategies. 🏦 Financial Services Implications → Model Risk: Competitive edge is found in the data moat, not the choice of LLM. → Settlement Risk: An AI hallucination constitutes a Compliance failure, not a clerical error. 3 Rules for Adoption ✅ Prioritize Auditing: A lack of data lineage precludes true ownership. ✅ Validate Competence: Technical proficiency does not equate to common-sense reasoning. ✅ The Kill Switch: Every agentic workflow requires a 60-second isolation protocol. 🗣️ C-Suite Inquiries → How do we develop 2030 leadership if entry-level roles are eliminated? → How do we address the $172B consumer value paradox?" → What is our 24-month 'Compute Independence' strategy? Robust governance enables high-velocity operations. Save this analysis for your next QBR. 📌 Data fueled by: Stanford University HAI, McKinsey & Company , LinkedIn , and Hugging Face . #AIGovernance #StanfordAIIndex #BoardroomStrategy #QuantumRisk #FinancialServices #NYAG #Gartner #ModelOp

I'm excited to announce that I'll be joining Bloomberg this summer in New York City as a Software Engineering Intern. Looking forward to building high-performance systems for global financial markets.

The Convergence: How ISO42001 and National Institute of Standards and Technology (NIST) AI RMF Work Together 🤝 You don't choose one; you Stack them. → Use NIST AI RMF to define technical risk appetite and "Red Teaming" protocols. → Use ISO - International Organization for Standardization 42001 to wrap those protocols into a certifiable, board-reportable management system. By 2027, the "Quantum Divide" will separate firms that can prove Post-Quantum Cryptography (PQC) readiness. As National Institute of Standards and Technology (NIST) finalizes PQC standards (FIPS 203/204/205), leading FS firms are already integrating these into their AI training pipelines to prevent future decryption of harvested data. Boardroom Command: 3 Questions for Senior Executives → The Containment Gap: "Do we have a 'Kill Switch' protocol that can isolate an autonomous AI agent from our sensitive core ledger in under 60 seconds?" → The Fiduciary Leak: "How are we quantifying the $547B global value leak against our specific AI investment portfolio?" → The Quantum Clock: "Are our AI training datasets being harvested today for future decryption, and what is our 24-month migration plan to PQC standards?" 2026 Stats 📊 → The Abandonment Rate: 60% of Enterprise AI projects will be abandoned by year-end 2026 due to unmanaged risk and poor data readiness (Source: Gartner ). →The Readiness Gap: Only 24% of Financial Institutions have integrated AI Governance into their existing Fiduciary Risk Frameworks (Source: ModelOp ). The 2026 Adoption Checklist ✅Map the Shadows: Inventory all "Shadow AI" using automated discovery. ✅Audit the Lineage: Implement evidence-quality audit trails; utilize the Cyber Risk Institute (CRI) FS AI RMF for sector-specific mapping. ✅Quantum-Proof the Core: Transition AI-integrated encryption to National Institute of Standards and Technology (NIST) approved post-quantum algorithms. The Call to Action Governance is not the brake; it is the reason you can safely go 200 MPH. This crosswalk is the baseline for 2026 resilience. As global leaders across Financial Services begin to codify these stacks, the distinction between "Adoption" and "Leadership" will be defined by Post-Quantum readiness. Save this guide for your next strategic review. 💾 #AIGovernance #NIST #ISO42001 #QuantumRisk #FinancialServices #BoardroomStrategy #NYAG #Gartner #ModelOp

Why Creating a Safe Space at Work Matters to Me 💗 | Day of Pink Creating a safe space at work matters deeply to me because everyone deserves to feel respected, heard, and supported—exactly as they are. On Day of Pink, I was reminded that safety isn’t just about policies or moments—it’s about everyday choices. It’s how we listen, how willing we are to learn from experiences different from our own, and how we reflect on the impact we have on others. That’s why the ECHO theme—Encouraging Listening, Learning, and Shared Reflection—resonates so strongly with me. I truly believe there is no place for bullying, exclusion, or silence in the face of disrespect. I support creating a work environment where people feel safe to speak up, be themselves, and thrive. Wearing pink this week is meaningful—but living these values every day is how we create real change. #DayOfPink #ECHO #EndBullying #SpeakUp #ListenLearnReflect #DEI #InclusionMatters #WorkplaceCulture #LeadWithEmpathy