cover_letter / art_W_MnaUiwRJo

Dear Stripe Payments Intelligence Hiring Team, Stripe's mission to increase the GDP of the internet is one of the few genuinely infrastructure-level ambitions in technology — the kind where the underlying platform determines what's even possible for millions of businesses. My interest in this role is grounded in a specific thread of my own work: at Intuit, I owned the platform that processed 675M+ engagements in FY23 across QuickBooks, TurboTax, Mint, Mailchimp, and Credit Karma, scaling throughput from 6K to 50K TPS via an rSocket migration supporting ~1.5M concurrent connections at sub-25ms TP99. That experience of building real-time infrastructure where latency and reliability are not aspirational but contractual maps directly to what Stripe's Payments Intelligence team does every day. ## Technical and AI/ML Foundation My technical foundation spans the full stack from ML research to production infrastructure. In 2004 I hand-coded backpropagation through time in C++ for a protein structure prediction system at UC Berkeley — work that led to a NeurIPS 2014 publication. In 2026 I rewrote that system as a production PyTorch platform spanning 413 parameters to 8 billion (a 19-million-fold scale increase), with MLflow experiment tracking, Optuna hyperparameter optimization, and FastAPI serving backed by 823 automated tests across six Docker containers. More recently, I built a post-training RL workbench covering the full RLHF/DPO pipeline: a Reward Lab for designing and A/B testing reward functions across GSM8K, MATH, HumanEval, and UltraFeedback; a Playground for real TRL-powered GRPO/DPO training with live SSE metric streaming on Apple Silicon and CUDA; and an Arena for head-to-head framework benchmarking across TRL, VeRL, OpenRLHF, and NeMo RL with GPU passthrough in Docker containers. I implemented 12 RL algorithms with standardized throughput, memory, and convergence benchmarking. I also built aeval, a local-first model evaluation platform with bootstrap confidence intervals, Welch's t-test, Cohen's d effect size, adversarial safety testing, and CI/CD regression detection — the kind of statistical rigor that matters when you are making real decisions about model behavior in production. On the fintech side, I built Fintellect AI — a mobile-first AI financial platform (iOS/Android/Web) — from prototype to App Store approval, architecting a RAG retrieval pipeline with ChromaDB, multi-provider LLM orchestration across Claude, GPT-4, and Gemini with fallback routing and token-budget optimization, and 13 specialized AI advisors delivering context-aware financial guidance. Building a consumer fintech product end-to-end — including navigating App Review, Apple IAP, App Store Server Notifications, and live Alpaca market data — gave me a practitioner's understanding of the financial data and trust requirements that Stripe's users depend on. ## Why This Role The Payments Intelligence team sits at the intersection of the two things I find most technically compelling: real-time ML serving at infrastructure scale and the product judgment required to translate model outputs into user-facing business impact. Owning products like Radar, Authorization Boost, and Disputes — where every prediction is part of a live transaction flow — requires exactly the combination of analytical depth, cross-functional coordination, and platform thinking that has defined my career. The opportunity to shape how ML controls are architected to optimize authorization rates and minimize fraud, at the scale Stripe operates, is the kind of problem I want to be working on. ## Selected Prior Experience - **Scaled Intuit's ICE platform to 675M+ engagements in FY23** across QuickBooks, TurboTax, Mint, Mailchimp, and Credit Karma; drove 275% YoY growth in engagements and scaled throughput from 6K to 50K TPS via rSocket migration supporting ~1.5M concurrent connections at sub-25ms TP99 — directly analogous to Stripe's real-time prediction serving requirements. - **Delivered ICE Self-Service platform** (DevPortal, GitOps config, ICE Playground), reducing developer onboarding from 2–3 weeks to minutes in pre-prod and under 24 hours for production, while mitigating $1M+ in projected opex growth — demonstrating the ability to translate platform investment into measurable business outcomes. - **Implemented ICE Presence in async chat**, generating $480K/month in additional invoicing; deployed Background-to-Foreground Messaging on iOS/Android with sub-100ms latency — evidence of shipping ML-adjacent features with direct revenue attribution. - **Built aeval** with statistical rigor (bootstrap confidence intervals, Welch's t-test, Cohen's d, saturation detection) and CI/CD safety gates — the evaluation discipline required to make trustworthy decisions about model behavior in a payments fraud and authorization context. - **Architected Fintellect's RAG pipeline** with multi-provider LLM orchestration, fallback routing, structured-output validation, and token-budget optimization — hands-on experience building production AI systems in a regulated financial domain. - **Owned Search Service, Search Catalog, and SPL/SPL2 at Splunk** — delivered Scheduler Service end-to-end in approximately four months and led a query performance initiative achieving up to 10x improvements for a Fortune 500 beta customer, demonstrating the ability to drive complex technical roadmaps with measurable results. - **Conducted enterprise-wide Service Language Assessment at Intuit** across nine languages, analyzing usage data and developer feedback to inform strategic investment decisions presented to the CTO — the kind of data-driven, cross-functional analytical work that Stripe's Payments Intelligence team requires when prioritizing across billions of transactions. ## Closing Stripe's goal of putting the global economy within everyone's reach is not a tagline — it is an infrastructure problem, and Payments Intelligence is one of the core mechanisms by which that infrastructure either works or fails for the businesses depending on it. I have spent my career at the intersection of platform scale, developer experience, and applied ML, and I am ready to bring that full arc to bear on the authorization, fraud, and revenue optimization challenges that define this team's mandate. I would welcome the opportunity to discuss how my background maps to the specific roadmap priorities on the Payments Intelligence team. Sincerely, **O. Felix Amoruwa** famoruwa@berkeley.edu · 909-731-9011 · felixamoruwa.info