How is AI reshaping the gamut of modern financial services? Mike Upchurch, VP of Strategy for Financial Services and Insurance, Domino Data Lab weighs in with some observations in this fintech interview:
______________
Hi Mike, take us through your fintech journey and the top observations you have for the market?
My journey spans from hands-on model development to strategic planning across retail, commercial, and investment banking. One observation is how AI mirrors the internet’s transformative impact. For example, the Internet democratized information and AI will democratize intelligence, they both fundamentally reshape business models, and create new competitive dynamics in ways that happen only once a generation.
What’s striking is the velocity of change, complexity, and risk management needs. Another observation is that organizations aren’t just adopting AI tools, they are rebuilding their operational foundations. It’s an interesting challenge to master the balance between innovation, operational discipline, and risk management.
How are you seeing different aspects of financial services being driven by AI powered tools today?
AI is fundamentally reshaping financial services through core enabling capabilities such as large language models (LLMs), tools like GitHub Copilot, and AI agents. These are driving innovations such as decentralized finance (DeFi) and the emergence of “cognitive banking,” where systems proactively anticipate and address customer needs. However, this transformative power also introduces significant considerations around regulatory compliance, brand reputation, and customer account security, necessitating robust governance and risk management frameworks.
How can modern AI powered fintech enable banks and finserv organizations with their unstructured data more optimally and what should banks and finservs keep in mind as they deploy AI powered tools to enhance and optimize processes?
LLMs have transformed 80% of enterprise data that was previously unusable. Take commercial lending, for example: AI can now extract key covenant terms from thousands of loan agreements, identify portfolio risk patterns, and flag potential defaults weeks earlier than traditional models.
Of course, every solution brings a few challenges. Unified, governed development environments help teams move fast without losing control I’d suggest three tactical priorities for deployment: first, establish data lineage tracking and curation so you know exactly what data feeds each model which reduces incorrect output and prevents issues like LLMs reading a version of a policy document that is no longer in force. Second, implement continuous model monitoring with automated drift detection, especially for models processing customer communications or data from sources that have changing levels of predictive power. Third, it’s critical for compliance to ensure reproducibility and that you can trace decisions to specific code, data, and environments.
The institutions succeeding at scale treat AI development like software engineering and implement version control, testing protocols, and deployment pipelines.
Read More: AI is Making Accounting and Finance Faster, Smarter and More Valuable
What should banks and financial services businesses who are still struggling with legacy systems consider when looking to streamline their digital transformation goals and processes?
Banks should layer modern tools over legacy systems using APIs so they get speed and reliability. For example, mainframes still outperform modern systems for high-volume transactions. Hybrid infrastructure allows selective and prioritized modernization through API-first integration layers. For instance, one regional bank kept their core deposit system but built modern APIs that enabled mobile banking innovation and real-time payment processing. The key is adopting a platform approach with standardized development environments, automated compliance checks, and controlled experimentation frameworks. This creates “innovation guardrails” so teams can move fast while staying within risk parameters.
Practically, start with customer-facing applications where legacy limitations are most visible, then work backward to identify the minimum viable modernization needed to unlock those capabilities.
Five thoughts on the future of fintech and the impact of AI on fintech before we wrap up?
The rapid evolution of AI, particularly with the explosion of GenAI, is fundamentally and rapidly reshaping financial services. Most firms can prototype GenAI. Few can scale it. Here are five key considerations for the future:
Embrace Orchestrated Complexity: Organizations that master managing AI’s inherent messiness with multiple models, diverse data sources, and evolving regulations will outcompete those seeking perfect order. Platform-based approaches provide structure without stifling innovation.
Build Model Factories: Institutions that can rapidly develop, test, and deploy models at scale will realize tremendous benefits. This requires treating model development like manufacturing and implementing standardized processes, quality control, and automated testing pipelines.
Embed Governance: AI governance must shift from end-stage gatekeeping to continuous, automated validation throughout the development lifecycle. This includes bias testing, explainability requirements, and regulatory compliance checks as integral parts of the workflow. MRM teams need to perform continuous validation, not just annual reviews. Excelling at governance is a secret weapon for scaling AI.
Organize Based on Hybrid Operating Models: Success requires C-suite leadership combined with federated execution teams. Centralized strategy and standards, combined with distributed implementation and innovation. This balance enables both agility and accountability.
Balance Human-AI Collaboration: The future isn’t about replacing human expertise but amplifying it. The most successful implementations will be those that enhance human decision-making rather than attempting to automate it entirely.
Read More: Global Fintech Interview with Nathan Shinn, Co-founder and Chief Strategy Officer of BillingPlatform
[To share your insights with us, please write to psen@itechseries.com ]
Domino Data Lab empowers the largest AI-driven enterprises and major government agencies to build and operate AI at scale. Domino’s Enterprise AI Platform provides an integrated experience encompassing model development, MLOps, collaboration, and governance. With Domino, global enterprises can develop better medicines, grow more productive crops, develop more competitive products, and more.
Mike Upchurch is the Vice President of Strategy for Financial Services at Domino Data Lab, bringing over 25 years of expertise in analytics, ML/AI, business strategy, and technology. Previously, Mike held roles at Capital One as a product manager in their innovation lab and as a strategy and operations consultant in their Center for Machine Learning. Mike led strategy at Notch and in the mortgage lending group of Bank of America and was the co-founder of Fuzzy
