cover_letter / art_MpyIR09pL6Y

Dear Anthropic Hiring Team, Anthropic's mission — building AI systems that are reliable, interpretable, and steerable — is not a tagline to me; it is the technical and ethical standard I have held myself to across 12 years of building developer-facing platforms, AI frameworks, and production ML systems. When I read about Project Glasswing, the KPMG and PwC deployments, and the Stainless acquisition signaling a serious developer platform investment, I see an organization at exactly the inflection point where deep technical credibility and partner-facing execution matter most. That intersection is where I have spent my career. **Technical Foundation** My AI/ML work spans from first principles to production scale. In 2004, I hand-coded a neural network in C++ with custom backpropagation-through-time for protein structure prediction at UC Berkeley — work that was accepted at NeurIPS 2014. In 2026, I rewrote that system as a full PyTorch platform spanning 413 parameters to 8B (a 19-million-fold scale increase), with MLflow experiment tracking, Optuna hyperparameter optimization, and FastAPI serving across five neural architectures. More directly relevant to Anthropic's alignment work, I built a 3-phase RL post-training workbench that benchmarks GRPO, DPO, PPO, DAPO, REINFORCE++, RLOO, SimPO, IPO, KTO, ORPO, and SPPO — 12 algorithms in total — across TRL, VeRL, OpenRLHF, and NeMo RL, with live SSE metric streaming on Apple Silicon (MPS) and CUDA, and GPU passthrough in Docker containers. I built reward function A/B testing across GSM8K, MATH, HumanEval, and UltraFeedback datasets. This is not familiarity with LLM frameworks; it is hands-on implementation of the post-training stack that produces models like Claude. On the multi-agent side, I designed and implemented 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/dental, and financial-markets verticals. I also built a RAG retrieval pipeline with ChromaDB, multi-provider LLM orchestration across Claude, GPT-4, and Gemini, fallback routing, structured-output validation, and token-budget optimization. These are the reference architectures that GSI and cloud partners are trying to build for their enterprise clients right now. **Why This Role** My arc — from ML researcher to platform infrastructure PM to founding AI companies that ship production systems — maps directly to what the Applied AI Architect, Partnerships role requires: the ability to sit with a GSI's engineering team, understand their deployment constraints, co-design a reference architecture, and then translate that into a joint solution that accelerates a customer deal. I have done each of those things, just not yet under the Anthropic banner. What specifically excites me about this role is the moment Anthropic is in. The KPMG deployment across 276,000 employees, the PwC enterprise build-out, and the new institutional enterprise AI services company with Blackstone and Goldman Sachs represent exactly the kind of large-scale, multi-stakeholder partner ecosystem where technical depth and GTM fluency have to coexist. The Stainless acquisition signals that Anthropic is investing seriously in the developer platform layer — an area where I spent three years at Intuit building SDK starter kits, DevPortals, and CI/CD scaffolding that reduced developer onboarding from weeks to minutes. I understand what it takes to enable a developer ecosystem at scale, and I understand the Claude API well enough to build production applications on top of it. **Selected Prior Experience** - **ICE Platform Scale (Intuit):** Achieved 275% YoY growth in ICE 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 latency — the kind of infrastructure credibility that earns trust with AWS and GCP partner architects. - **Developer Enablement (Intuit):** Delivered the 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; extended Java and Python SDK Starter Kits with scaffolding templates, build configurations, testing frameworks, and CI/CD integration. - **Enterprise Language & Platform Strategy (Intuit):** 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 cross-organizational, C-suite-facing technical communication this role requires. - **RL Post-Training Workbench:** Built and published a 3-phase workbench benchmarking 12 RL algorithms across TRL, VeRL, OpenRLHF, and NeMo RL with live metric streaming — directly relevant to enabling GSI AI practices around fine-tuning and post-training workflows. - **OpenClaw Multi-Agent Framework:** Designed gateway protocol, subagent delegation, and session management for cross-industry AI agent orchestration — a reference architecture directly applicable to the agentic deployment patterns Anthropic's enterprise partners are building. - **Fintellect AI — Agents for Financial Services:** Architected a production RAG pipeline with multi-provider LLM orchestration, 13 specialized AI advisors, and a Mac App Store–sandboxed trading copilot — aligning with Anthropic's vertical push into financial services agents. - **Search Orchestration (Splunk):** Owned Search Service (Go microservices), Search Catalog (PostgreSQL metadata), and SPL/SPL2; delivered Scheduler Service end-to-end in ~4 months; led query performance optimization achieving up to 10x improvements for a Fortune 500 beta customer — demonstrating the ability to own complex technical programs from design through delivery. - **Conference & Community Presence:** Speaker at DeveloperWeek 2022 and Splunk .conf18/.conf19 — experience representing technical platforms in front of developer communities, which maps directly to the GSI workshops, AWS Summits, and hackathons this role requires. **Closing** Anthropic's commitment to building AI that is safe and beneficial is not in tension with shipping capable, commercially successful systems — it is the foundation that makes enterprise partners like KPMG, PwC, and Goldman Sachs willing to build on Claude. I want to be the technical bridge that makes those partnerships work: enabling GSI AI practices, co-designing reference architectures, unblocking strategic deals, and feeding ground-truth deployment feedback back into Anthropic's product roadmap. I have the ML depth to earn credibility with research teams, the platform PM experience to design scalable solutions, and the founder's instinct to move fast when a partner needs to ship. I would welcome the opportunity to discuss how my background maps to Anthropic's partnership priorities. Sincerely, **O. Felix Amoruwa** famoruwa@berkeley.edu | 909-731-9011 | felixamoruwa.info