Adi CEO Reveals AI-First Fintech Strategy in Latin America
Adi CEO Santiago Suarez shares how contrarian market entry, monorepo architecture, and AI-native operations drive fintech scale in Colombia. Learn actionable insights on building technology-first companies, deploying AI agents, and optimizing organizational metrics for profitability.
Adi demonstrates how contrarian market selection and foundational tech architecture drive fintech dominance in emerging markets. CEO Santiago Suarez reveals that Adi's success stems from rejecting conventional wisdom to target Colombia, a market characterized by fragmented financial infrastructure, cash dominance, and high smartphone penetration. This focus allowed Adi to build a comprehensive platform spanning BNPL, payments, logistics, and banking, serving over 3 million consumers and 50,000 merchants while avoiding the saturation of larger markets like Brazil. Suarez emphasizes that consensus plays yield no alpha, urging entrepreneurs to pursue ambitious goals through unconventional paths. He also highlights the importance of learning from peers like Kaspi, adopting lessons on modular product roadmaps and ignoring short-term equity investor pressure to maintain long-term strategic focus.
AI-Native Architecture and Data Strategy
Adi's competitive edge relies on deliberate technical decisions made years in advance. The company built on a monorepo architecture rather than microservices, a choice that significantly streamlined code accessibility for large language models and reduced duplication. Coupled with an event-sourcing system logging over 10 million daily events via Kafka and Databricks, Adi created a unified data foundation. This infrastructure supports 200 production agents, including customer service bots achieving 80% resolution rates and merchant onboarding tools that improved conversion by over 20%. Suarez advises launching AI initiatives with complex, high-stakes workflows like legal compliance to force the development of robust data pipelines before scaling to simpler use cases. The company is also developing its own transformer model, ADDNA, to further customize AI capabilities. This approach ensures that AI integration is not a wrapper but a core component of the engineering stack.
Operational Discipline and Talent Acquisition
Operational rigor underpins Adi's efficiency. The company runs in English to attract global talent and elevate local standards, while a remote-first, written culture ensures all processes are documented for AI context. Suarez replaced cascading OKRs with a single North Star metric focused on profitability, sharpening organizational focus during market downturns. This discipline has enabled Adi to run 150 headcount positions below budget while exceeding growth targets, proving that AI-driven automation and explicit standards can decouple revenue growth from linear cost increases. Adi's model offers a replicable framework for entrepreneurs seeking to build scalable, technology-first enterprises in underserved regions by prioritizing deep tech investment, modular thinking, and relentless focus on customer pain points. The emphasis on explicit documentation transforms tacit knowledge into actionable data, creating a self-reinforcing loop where human processes train agents, and agents augment human productivity. Suarez warns against moving too fast initially, citing early fraud risks, but advocates for aggressive scaling once foundations are secure. The combination of monorepo efficiency, event-driven data, and AI agents positions Adi to expand its footprint across Latin America while maintaining superior unit economics.
Key insights
-
Adi's monorepo architecture, chosen against the microservices trend, significantly reduced code duplication and enabled seamless LLM integration, accelerating AI agent deployment.
Impact: Companies adopting monorepos can achieve faster AI adoption cycles and lower maintenance costs compared to fragmented microservice environments.
-
Starting AI implementation with high-stakes legal workflows forced the creation of robust data pipelines and feedback loops, which then scaled effectively to customer service and onboarding.
Impact: Prioritizing complex, high-risk AI use cases builds stronger infrastructure and ensures reliability before expanding to broader applications.
-
Adi's focus on Colombia, a market with high smartphone adoption but fragmented financial infrastructure, allowed the company to capture significant share without competing in saturated markets.
Impact: Entrepreneurs can find alpha by targeting underserved markets with unique structural inefficiencies rather than following consensus plays in larger economies.
-
A rigorous written culture of memos and SOPs provides the explicit context required for AI agents to operate autonomously, reducing dependency on tacit knowledge.
Impact: Organizations that document processes thoroughly can leverage AI to automate workflows more effectively and scale operations without linear headcount growth.
-
Replacing complex OKR cascades with a single North Star metric focused on profitability sharpened organizational alignment and drove efficiency during capital-constrained periods.
Impact: Simplifying performance metrics to a single focus area can improve execution speed and ensure all teams contribute directly to core business objectives.
Action items
-
Audit current architecture for monorepo viability to improve code accessibility for AI agents and reduce duplication across engineering teams.
Impact: Transitioning to a monorepo can accelerate AI integration and lower long-term maintenance costs by centralizing code management.
-
Implement a comprehensive documentation strategy for all SOPs and workflows to create structured context for AI agent deployment.
Impact: Explicit documentation enables AI to automate complex processes reliably, reducing reliance on human intervention and scaling operations efficiently.
-
Launch AI initiatives with high-complexity, high-stakes workflows to build robust data pipelines and feedback loops before scaling to simpler use cases.
Impact: Starting with difficult problems ensures the AI infrastructure is resilient and capable of handling nuanced scenarios across the organization.
-
Consolidate performance metrics into a single North Star metric aligned with profitability to eliminate distraction and focus organizational efforts.
Impact: A unified metric simplifies decision-making and ensures all teams prioritize actions that directly impact the company's financial health.
Quotes
“Don't let your ambition fall prey of commercial wisdom. If you're a consensus play, there's just no alpha.”
“NPS for a company with less than a thousand employees is a terrible metric. You should just listen to customer service calls.”
“Remember you're a technology company... that's actually where a lot of your equity value eventually compounds.”