cover_letter / art_NR6wfDi-Uuc

Dear Stripe Global Transformation Team, Stripe's mission — increasing the GDP of the internet — is not an abstraction. It is the infrastructure layer that determines whether a founder in Lagos or a startup in Seoul can participate in the global economy on equal footing. That framing resonates with me directly: I spent the last three years building Fintellect AI and Streamio AI from zero, navigating App Store approvals, live trading infrastructure, and multi-agent orchestration, precisely because I believe financial access and AI tooling should be available to everyone, not just those with enterprise contracts. When I read that the Global Transformation Team's mandate is to make ProServ AI-native and prove a model that becomes the blueprint for how Stripe transforms itself everywhere, I recognized the same 0-to-1 product challenge I have been living. --- **Technical Foundation** My technical credibility is grounded in building AI systems end-to-end, not just managing teams that do. In 2026 I shipped a full RL post-training workbench covering the complete RLHF/DPO pipeline: a Reward Lab for designing and A/B testing reward functions across GSM8K, MATH, HumanEval, and UltraFeedback; a Playground running real TRL-powered GRPO and 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. I implemented 12 RL algorithms — PPO, GRPO, DAPO, DPO, SimPO, and others — with standardized throughput, memory, and convergence benchmarking. This is not PM-adjacent familiarity with AI; it is the kind of depth that lets me hold a substantive engineering conversation about why a reward shaping choice matters or what a convergence curve is actually telling you. On the agentic side, I designed and shipped OpenClaw, a multi-agent orchestration framework with a gateway protocol, subagent delegation, profile management, and session switching — coordinating AI agent workflows across real estate, insurance, health, and financial-markets domains. I also built a RAG retrieval pipeline with multi-provider LLM orchestration (Claude, GPT-4, Gemini), fallback routing, structured-output validation, and token-budget optimization for Fintellect AI. These are the exact categories — agentic AI, LLMs, workflow automation — that the JD identifies as the relevant landscape for this role. At Intuit, I operated at a different scale: 675M+ ICE engagements in FY23 across QuickBooks, TurboTax, Mint, Mailchimp, and Credit Karma, with throughput scaled from 6K to 50K TPS via rSocket migration supporting approximately 1.5M concurrent connections at sub-25ms TP99. That experience taught me what it means to ship platform infrastructure that real engineering teams depend on globally — and how to earn the trust of engineers who will reject a PM who cannot speak their language. --- **Why This Role** The arc from hand-coding BPTT in C++ at Berkeley in 2004, to publishing at NeurIPS, to building production AI products as a founder, to now wanting to operate inside a company with Stripe's infrastructure reach — that arc is not accidental. I am at a point where I want the leverage of Stripe's scale behind the AI transformation work I know how to do. The ProServ AI PM role sits exactly at the intersection I find most interesting: AI strategy, engineering execution, and organizational change, with a global remit and direct access to senior leadership. What specifically draws me to this role is the mandate to coordinate a distributed network of engineers and consultants who do not report to you, structure their contributions into well-scoped workstreams, and drive adoption globally. That is a description of the coordination problem I solved at Intuit — influencing across 20+ mobile apps and 30+ product SKUs without direct authority over every engineering team — and the same muscle I used as a founder managing external API partners, contractor engineers, and App Store review constraints simultaneously. The JD's emphasis on defaulting to action and shipping something imperfect to iterate resonates with how I actually work: Vantage, Fintellect, and Streamio were all shipped iteratively under real constraints, not after perfect alignment. --- **Selected Prior Experience Most Relevant to This Role** - **Intuit ICE Platform:** 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 — a transformation program with measurable adoption outcomes. - **Intuit Scale & Telemetry:** Worked closely with telemetry and usage data (SQL, BigQuery) to prioritize developer pain points across ~20 mobile apps and 30+ product SKUs; used data to drive roadmap decisions presented to the CTO. - **Intuit Cross-Geo Influence:** Led Mailchimp GCP-to-AWS migration for MSaaS and conducted enterprise-wide Service Language Assessment across 9 languages — both required influencing engineering teams across organizational boundaries without direct authority. - **OpenClaw Multi-Agent Orchestration:** Designed and implemented a multi-agent gateway framework coordinating AI workflows across four industry verticals — directly applicable to building AI tooling that ProServ consultants can embed into client engagements. - **RL Workbench & aeval:** Built production AI evaluation infrastructure (aeval: FastAPI, TimescaleDB, Redis, Ollama; RL Workbench: 12 algorithms, cross-framework benchmarking) — establishes the technical credibility to govern AI initiatives and communicate trade-offs to an AI SteerCo with precision. - **Splunk Scheduler Service:** Delivered Scheduler Service end-to-end in approximately 4 months, enabling scheduled search capabilities for first-party applications and demoed at .conf19 — evidence of moving with urgency from roadmap to shipped product. - **De Anza College Adjunct Faculty:** Teaching cloud computing, data analytics, and Java programming since 2018 — a consistent track record of translating complex technical material into accessible frameworks for diverse audiences, which maps directly to developing AI playbooks for ProServ teams. --- **Closing** Stripe's stated goal is to put the global economy within everyone's reach. The Global Transformation Team's goal is to make ProServ AI-native so that Stripe's users realize the full value of the platform faster. Those two goals are connected: the quality of the professional services delivery layer determines whether enterprise customers — and the businesses they serve — actually unlock what Stripe makes possible. I want to be the person who builds that AI-native ProServ model, proves it works globally, and hands the blueprint to the rest of the organization. I would welcome the opportunity to discuss how my background maps to what you are building. Sincerely, **O. Felix Amoruwa** famoruwa@berkeley.edu · 909-731-9011 · felixamoruwa.info