cover_letter / art_F0N6Gf18JvQ

Dear Brex Hiring Team, Brex is building what a modern financial OS should look like — not a patchwork of legacy tools, but an AI-native platform where spend management, corporate cards, and banking converge into a single, intelligent system. That mission resonates with me directly: I spent the last year building Fintellect AI, a mobile-first platform that merges real-time market insights, RAG-powered financial agents, and multi-provider LLM orchestration to help retail investors make better financial decisions. The problem space — using AI to reduce friction and improve decision quality in financial workflows — is one I've been living inside. **Technical and AI Foundation** My AI/ML work spans from foundational research to production systems. My NeurIPS 2014 paper on neural networks for protein structure prediction was built on a system I originally hand-coded in C++ with custom backpropagation through time in 2004 — a 2026 rewrite now spans 413 parameters to 8B, a 19-million-fold scale increase, with PyTorch, MLflow, Optuna HPO, and FastAPI serving across 6 Docker containers. That arc — from first principles to production infrastructure — shapes how I think about AI product development. More recently, I built an RL post-training workbench that benchmarks GRPO, DPO, PPO, DAPO, and eight other algorithms across TRL, VeRL, OpenRLHF, and NeMo RL frameworks, with live SSE metric streaming, GPU Docker passthrough, and cross-tab workflow lineage tracking. I also built aeval, a local-first model evaluation platform with bootstrap confidence intervals, Welch's t-test, Cohen's d effect sizing, and automated safety gates — the kind of statistical rigor that separates genuine AI capability measurement from demo theater. These aren't research prototypes; they're the tools I use to make product decisions about AI systems. At Fintellect AI, I architected a RAG retrieval pipeline with ChromaDB, multi-provider LLM orchestration across Claude, GPT-4, and Gemini with fallback routing, structured output validation, and token budget optimization. At Streamio AI, I implemented the OpenClaw multi-agent orchestration framework with gateway protocol, subagent delegation, and session switching — coordinating AI agent workflows across financial, real estate, and insurance verticals. **The Bridge** The through-line in my career is building platforms that developers and end users depend on at scale — and doing it with enough technical depth to make the right architecture calls, not just manage around them. Brex's AI-native automation ambition, combined with its enterprise push and the complexity of coordinating across Engineering, Risk, Legal, Compliance, and GTM, is exactly the operating environment where that combination matters. **Why This Role** What draws me to the Group PM role specifically is the scope: owning a multi-quarter vision for a core product area, developing a team of PMs, and driving cross-functional alignment on initiatives that don't fit neatly into one team's roadmap. Brex AI — expense categorization, policy enforcement, financial insights — sits at the intersection of LLM capability and financial workflow automation, which is precisely the product surface I've been building toward. The enterprise push adds another dimension I find compelling: the prioritization tradeoffs between startup agility and Fortune 500 compliance requirements are non-trivial, and navigating them requires the kind of data-driven, stakeholder-fluent operating model I've developed across Intuit, Splunk, and Kaiser Permanente. **Selected Prior Experience** - At Intuit, achieved 275% YoY growth in ICE platform engagements, scaling to 675M+ in FY23 across QuickBooks, TurboTax, Mint, Mailchimp, and Credit Karma; scaled throughput from 6K to 50K TPS via rSocket migration supporting ~1.5M concurrent connections with sub-25ms TP99. - 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. - Implemented ICE Presence in async chat, generating $480K/month in additional invoicing — a direct example of connecting platform infrastructure investment to measurable revenue outcomes. - Conducted enterprise-wide Service Language Assessment across 9 languages, analyzing usage data and developer feedback to inform strategic investment decisions presented to the CTO — the kind of data-driven, executive-facing prioritization work this role requires. - At Splunk, led query performance optimization for a beta customer, building a mirrored Enterprise topology for benchmark testing and achieving up to 10x performance improvements in Splunk Cloud Services search. - Designed a repeatable RICE-based prioritization framework for 3 microservice backlogs, balancing internal partner, third-party developer, and Fortune 500 customer requirements simultaneously. - At Fintellect AI, built domain-specific conversational AI agents scoped to distinct financial focal points, delivering guided, context-aware advisory interactions — a direct analog to Brex AI's expense and policy automation surface. **Closing** Brex's mission — enabling companies to spend smarter and move faster — is one I want to help execute at scale. The combination of AI-native product thinking, fintech domain depth, platform infrastructure experience, and the ability to develop PM teams and drive cross-functional alignment is what I bring. I'd welcome the opportunity to discuss how my background maps to Brex's current priorities. Sincerely, **O. Felix Amoruwa** famoruwa@berkeley.edu | 909-731-9011 | felixamoruwa.info