Disrupting Portfolio Management with AI
By Federico Kunze Küllmer, Founder @ Moneda
TL;DR
Users want hybrid finance: Human advice for complex decisions, AI for routine portfolio tasks. Not a contradiction, a clear signal.
Banks missed the mark: Traditional and neobanks failed to blend trust and automation. Younger users moved on.
Self-serve + HITL is the future: AI agents handle execution; humans provide reassurance. Better UX, full control, and regulatory alignment.
AI-managed portfolios will explode: From 5% today to a majority share, as agents gain context-awareness and RWAs move onchain.
Moneda is building this now: A 24/7 AI financial co-pilot with human-in-the-loop oversight, DeFi access, and personalised, programmable portfolio automation.
Trust Scales Poorly, AI Doesn’t
Trust remains a cornerstone of banking and financial relationships. This isn’t at all surprising, as financial matters are private, delicate, and crucial. In surveys, this becomes very clear. 71% of clients say they prefer having a dedicated relationship manager, someone they can confide in and trust with their financial matters. Yet at the same time, 65% of users prefer to manage most portfolio tasks digitally, and what’s more, 55% of Europeans already use robo-advisors on a monthly basis, instead of human counterparts. Why? Because convenience matters.
What users want is not a contradiction, but a clear signal: they value human advice and a personalised approach for complex decisions, but prefer digital interfaces for day to day activities. In other words, people want fast, automated execution for simple tasks, and human reassurance only when decisions feel more important and complex.
Why Banks (and Neobanks) Missed the Hybrid Curve
Traditional financial institutions never truly bridged the gap between human trust and digital convenience. Their systems were either high-touch but high-friction or digital but devoid of intelligence and personalisation. They couldn’t scale human relationships to meet modern expectations, nor could they embed meaningful financial guidance into their digital channels.
As a result, they failed to deliver a cohesive hybrid model; one that offers both strategic human support and smart automation. Consumers moved on. Younger investors in particular have reshaped the market: 72% of Gen Z and Millennials now hold most of their wealth in passive ETFs, and 61% use robo-advisors.
The message is clear: people don’t want a banker in their pocket. They want an intelligent financial assistant that works 24/7, understands their goals and preferences, and earns their trust without wasting their time.
Why Self-Serve and ‘Human‑in‑the‑Loop’ (HITL) Will Define the Next Decade in Finance
Imagine your portfolio running on autopilot, with oversight. A 24/7 AI financial agent continuously monitors markets, gives insights on your spending, and learns from your past activity. As this agent detects new opportunities in the market, it generates pre-built, personalised transactions to optimise your portfolio allocation together with its reasoning. Each recommendation prompts a push notification on your phone that you simply review and sign using your passkey or biometrics.
No spreadsheets. No phone calls. No branch appointments. Just review, approve, done. Simple. I call this the self-serve model.
But self-serve doesn’t mean solo. It can be turned into a hybrid model through an in-app integration that allows users to message a (human) financial analyst for additional financial support. What's notable is that architecture is already being used for customer service, where AI agents are integrated via Model Context Protocols (MCPs) to internal and third-party APIs that provide context about the user, their activity, and past conversations to the support agents (human or AI). The result is more than just a helpful chatbot. A hybrid human-in-the-loop (HITL) model for personal banking or finance not only provides the human point of contact that customers need to build trust in a financial brand but also dramatically improves the user experience by allowing the AI agent to build knowledge about your habits and preferences.
What about compliance, though? This self-serve model fits the limited‑risk tier (i.e, the second-lowest category) of the EU AI Act. Why? Because the AI only recommends. Whereas the user (human) remains the final decision‑maker by signing the transaction with their passkey after reviewing it. This provides full transparency to the user whilst giving them full sovereignty over the final decision. Happy user, happy regulator.
Looking ahead, we at Moneda expect regulators to permit autonomous execution within user-defined parameters; for instance, up to a capped amount per day or month, or within an approved risk tolerance. In this model, the user (not the institution or the regulator) sets the limits, effectively defining the operating envelope of their financial agent using prepay or top-up mechanics. That’s what responsible AI-native banking looks like: smart automation, embedded human judgment, and full user control by design.
Up Only: AI Projections in Finance
Only 5% of the global AUM is managed by AI or automated robo-advisors today. If we follow the trends in the development of AI models, we can expect that agents will start becoming embedded in finance within the current decade. As as result, there will be a significant increase in the number of customer portfolios that will be fully allocated by specialised financial AI agents. These agents are going to have context awareness and understand the user’s financial situation and preferences, just like portfolio managers today do, but at a global scale. The share of global AUM managed by AI will only boom.
Additionally, as stablecoins and tokenized RWAs move to massive adoption, a greater number of portfolios will be moved onchain because of their convenience, 24/7 markets, and transparency. Analysts project up to $16T in tokenised equities, bonds, and real estate by 2030. AI has a unique opportunity here, as agents can access deeper liquidity from decentralised finance (DeFi) and execute transactions with near instant execution. That’s what we’re shipping.
Moneda’s AI Agent
Users increasingly rely on algorithmic tools for round-the-clock monitoring, rebalancing, and low-cost execution, yet their willingness to act still hinges on the judgement and empathy of a human adviser. Fintechs that seamlessly weave these capabilities (autonomous AI for the grunt work, humans for strategic sign-off) will define the next decade of portfolio advice, because they simultaneously satisfy users’ demand for convenience, trust, personalisation, and regulatory transparency.
At Moneda, we believe that by combining the best aspects of portfolio advise from traditional banking with the digital interface of neobanks, we can improve the world’s financial literacy rates, democratise relationship management, and help users grow their wealth. We’re building a personal AI financial agent that:
Has direct access to your past activity: banking and stablecoin transactions, card purchases and investments.
Gives you personalised insights into your spending habits and provides recommendations to create better financial habits.
Monitors your portfolio and markets via in-house and third-party MCP integrations.
Learns about your risk profile in real time based on your preferences, behaviour, and activity.
Acquires assets for your portfolio by directly accessing a virtual card.
Allocates funds across crypto, RWAs (tokenised stocks and T-bills), and DeFi yield with a single instruction to build your portfolio.
Provides portfolio rebalancing recommendations for you to review and sign off.
Creates a transaction from plain text for you to transfer EUR or USD stablecoins to any European bank or major blockchain network (EVMs, Solana, Bitcoin, etc).
The result? 24/7 institutional‑grade portfolio management for everyone. A personal co-pilot to help you navigate your finances and grow your wealth. An exponential improvement in the banking experience.
We're launching Moneda’s private beta soon. If you’re interested in how programmable money and AI can reshape personal finance, join the waitlist and help us test what comes next.
👉 Join the Waitlist: https://moneda.typeform.com/waitlist?utm_source=linkedin
Follow me and Moneda on LinkedIn for more insights
#Fintech #AI #DeFi #RWA #Tokenization #Stablecoins #Banking #Finance #Porfolio #Investing #DeepTech