Financial services are moving from rule-based software to continuous, adaptive systems that operate in real time. The biggest opportunities we see are not incremental UX improvements, but platforms that move us from insight to action, and fundamentally change how financial decisions are made and executed.

We’re particularly excited about:

• Continuous compliance and fraud prevention
  
  AI enables compliance to move from periodic audits to always-on systems that map regulation directly to live operations. These platforms answer questions like “where are we exposed right now, and what should change?” and increasingly take action themselves. The same shift is happening in fraud, where static rules are being replaced by adaptive, behavioural systems.

• Modern financial infrastructure
  
  Beneath the surface, we’re excited by new core banking systems, payment rails, treasury tooling, and identity layers that are built for programmability, automation, and cross-border use from day one - the pipes that make autonomous finance possible.

• AI-native wealth management
  
  The marginal cost of personalised financial advice is collapsing. This unlocks wealth tools for people previously ignored by traditional advisors, while enabling agentic systems that move money automatically based on user goals, constraints, and risk preferences. Over time, wealth apps stop being dashboards and become delegates.

What we think wins: products that embed directly into financial workflows, own the decision loop (not just the interface), and compound advantage through data and switching costs. 

The biggest change in work isn’t remote vs. office, it’s that execution is increasingly done by machines, not people. This creates a new category of infrastructure: systems that manage, coordinate, and audit non-human labour. Executed well, it could finally herald the end of years of flat productivity growth in the UK and beyond.

Key areas of conviction:

• Orchestration layers for AI labour
  
  As organisations deploy dozens (soon hundreds) of agents, orchestration becomes mission-critical. These platforms route tasks, enforce permissions, monitor outcomes, and provide auditability, effectively replacing layers of middle management. The winners here will look more like operating systems than tools.

• Vertical AI that refocuses entire job functions on what matters
  
  In industries like real estate, construction, legal, accounting, and infrastructure, AI is now capable of expert-level output. The opportunity isn’t productivity software, it’s end-to-end systems that do the work, compressing timelines from months to days and dismantling legacy cost structures.

• AI in the physical world
  
  We’re seeing a shift from digital-only applications to AI embedded in physical systems: predictive maintenance, asset intelligence, infrastructure monitoring, and robotics. These companies benefit from natural moats including proprietary sensor data, regulatory lock-in, and safety-critical workflows.

What we think wins: platforms that become the system of record for decisions and execution, accumulate proprietary data through use, and are difficult to displace once embedded. 

Consumer products are entering a new phase. We want to see the next generation of products that take work off a user’s plate so they can focus on genuine connection and fulfillment, feeling like personal agents.

We’re especially interested in:

• Personal health companions
  
  Rather than point solutions (fitness, sleep, supplements), we’re excited by systems that sit above wearables, calendars, labs, and lifestyle data, and translate signals into action. The defensibility comes from deep personalisation and owning the feedback loop between life and biology.

• Life orchestration tools
  
  From travel to scheduling to personal admin, we’re seeing consumer products move up the stack from tracking and insight to coordination and execution. These products reduce cognitive load by becoming the place where ‘what should happen next’ is decided.

• Agentic dating and relationships
  
  Dating is emotionally and cognitively exhausting. We’re seeing products emerge that help users understand their communication patterns and what they’re looking for in a partner - acting more like a supportive concierge than a matchmaking engine, with the potential to expand into broader relationship support over time.

What we think wins: products that earn trust, learn continuously, and become harder to replace the longer you use them-not because of habit, but because they genuinely understand you.