Strategic AI Adoption for Asset Managers and Enterprise Decision Makers - with Robert Kubin of Amundi

Robert outlines three characteristics making asset management ideal for AI: highly scalable service-based business model, mature competitive landscape with pricing constraints due to regulation, and inability to increase fees forcing firms to scale through technology investments rather than price increases.

• •Asset managers cannot increase product prices due to regulatory constraints and competition, forcing growth through scaling
• •Industry leaders invest heavily in technology but see limited productivity gains due to legacy system patching
• •Service-based model allows for easy scalability once proper AI infrastructure is in place
• •Regulatory approval required for fee increases creates unique pressure to optimize operations

Discussion of how traditional technology investments in asset management result in fragmented, non-communicating systems that quickly become obsolete. The conversation introduces the concept of 'stilt technology' versus comprehensive modernization, highlighting how firms end up with separate systems for decision-making, HR, finance, and NAV calculation that don't integrate.

• •Asset managers typically have multiple disconnected systems (HR, finance, NAV calculation, decision-making) that don't communicate
• •Technology upgrades become obsolete quickly due to rapid advancement, creating continuous investment cycles
• •Strict headcount policies combined with manual processes create friction and operational deficiencies
• •AI offers two primary value propositions: improved decision-making and process automation/efficiency gains

Robert explains the critical first steps for AI adoption, emphasizing that C-suite buy-in is deceptively difficult despite sounding simple. He outlines a practical organizational structure involving dedicated AI departments, local AI champions across geographies and business lines, and the importance of making AI part of organizational DNA rather than a separate initiative.

• •C-suite may resist due to unfamiliarity with technology; middle management too busy with daily tasks to innovate
• •Establish central AI department to evaluate use cases and coordinate strategy across organization
• •Deploy local AI champions for each geography, business line, and function to drive adoption
• •Build versus buy decision depends on organization size: larger firms should build more internally, smaller firms should acquire solutions
• •Small asset managers can now compete at scale of large firms by purchasing AI capabilities

Robert provides concrete examples of process improvement applications in asset management, focusing on KYC (Know Your Customer) onboarding and marketing materials generation. These use cases demonstrate how AI handles unstructured data from multiple sources to automate time-intensive manual processes.

• •KYC process involves aggregating data from external databases, client emails, and various formats - AI can structure and validate this automatically
• •Marketing materials require synthesizing market data, geopolitics, product information, and fund documentation on monthly/quarterly basis
• •Start with small portions of manual processes, test AI solutions, then scale successful implementations
• •Process improvement is easier starting point than decision-making enhancement for initial AI deployment
• •AI excels at extracting relevant information from unstructured, multi-source data sets

Discussion of practical vendor evaluation criteria for AI procurement, including technology validation, use case analysis, customer references, and pilot testing. Robert emphasizes treating AI vendor selection like any other enterprise procurement process with structured evaluation phases.

• •Evaluate 10 providers, shortlist to 3 based on technology capabilities, use cases, and customer experience
• •Request sample improvements on small processes before committing to larger implementations
• •Validate that results justify costs through pilot programs
• •Leverage technology teams to assess technical compatibility and integration requirements
• •Review vendor's customer base and industry-specific experience in asset management

Robert addresses the challenge of demonstrating ROI to quarterly-focused C-suites by providing concrete quantification methods. He outlines how to calculate cost reductions, headcount savings, and time efficiency gains to build business cases for AI investments.

• •Calculate direct cost decreases from process improvements and automation
• •Quantify headcount savings: if expansion requires 100 people without AI but only 50 with AI, calculate the difference
• •Measure time reduction: tasks taking 2 days reduced to 30 minutes translate to quantifiable cost savings
• •While exact outcomes aren't known upfront, ROI can be calculated with reasonable accuracy
• •Focus on early wins to demonstrate value while building toward broader strategic vision