Jul 31, 2025, 10:01 AM ETFireside Chat, Interview

The Finance Startup Bringing Agentic AI to Wall Street

Contract Velocity: In the last seven days, Model ML signed the same number of contracts as the entire fourth quarter, marking a shift from the previous year's "testing" phase to actual deployment.
Adoption Scope: Approximately 10% of the world's largest private equity firms, investment banks, asset managers, sovereign wealth funds, and venture firms now use the platform.
Sales Cycle Shift: High-level financial institutions have transitioned from buying software every decade to purchasing annually; average contracts are now multi-year, driven by immediate tangible value.
Decision Makers: Deals are approved by C-suite executives (CEOs/Executives) who place this as a top strategic agenda item, rather than mid-level operational teams.

Core Function: Model ML builds an "agentic" AI workspace for financial services, effectively replicating a cognitive architecture that mirrors a human's digital access to files, emails, CRMs, data vendors, and public filings.
User Interface: The system overlays this agentic brain onto familiar interfaces (Excel, Word, PowerPoint) to minimize user friction and adoption barriers.
Automation Capabilities: The platform automates the creation of complex documents (e.g., earnings summary slides, investment one-pagers) by pulling data from multiple sources, reducing processes that previously took days to hours or minutes.
Accuracy: Agentic systems currently demonstrate higher accuracy than humans in low-level data gathering and structured data extraction from public filings, though human oversight remains for confidence.
Technological Evolution: The industry has shifted from "testing" models to "using" them, driven by improvements in function calling, vision models (OCR and chart analysis), and autonomous task execution.

Work Ethic & Hiring: The founders emphasize "perseverance" over blind persistence; they currently work six to seven days a week and prioritize cultural fit and the question "Do you want to spend time with this person?" over CV prestige.
Hiring Unlearning: Former finance and quantitative employees must unlearn "big company thinking" (e.g., creating slide decks) in favor of first-principles thinking and rapid execution ("build the plane while flying").
YC Culture: The founders credit YC for instilling a weekly/daily operational rhythm and providing an instant support network that cannot be replicated in Europe.
Co-Founder Dynamic: As siblings, the founders maintain high transparency with no filter, citing "founder fallout" as the primary startup risk; they leverage a Venn diagram strategy where Arne handles engineering/product and Chas handles finance/commerce, with a large overlap in customer interaction.
Previous Ventures: Both founders previously sold YC companies: Fat Llama (a peer-to-peer rental marketplace with insurance) and Fancy (a vertically integrated grocery delivery service acquired by GoPuff).

Global Presence: The team is distributed across London, New York, Hong Kong, Singapore, and recently opened an office in India to support global customer base and engineering needs.
Customer Acquisition: The sales process relies heavily on in-person demos with real data to build trust, as clients fear professional risk; 80% of customers are US-based.
Location Debate: The founders argue that while London has strong talent, San Francisco offers a unique "buzz," higher density of ambitious peers, and better access to decision-makers, recommending founders move to tier-one cities or SF if possible.
Future Outlook: The founders predict a shift where user interfaces become less relevant as tasks become fully autonomous, occurring before the user even logs in ("things will already be done").

Motivation: The founders are driven by the impact of making people's lives easier (e.g., the "priceless" moment of a student receiving a delivery instantly or a financial analyst getting an immediate report) rather than just revenue.
Advice for Early 20s: Founders advise aspiring entrepreneurs to "go all in" if they possess passion and perseverance, noting that the "worst case" is simply gaining rapid, valuable experience and that the opportunity cost of a job is lower than the learning curve of a startup.
Product Market Fit: While Fancy found fit overnight due to timing (pre-COVID), Fat Llama took three years of iteration to prove unit economics and solve trust issues, highlighting the necessity of endurance.