cover_letter / art_1Pyu-_f6Nb8

Dear CoreWeave Developer Experience Hiring Team, CoreWeave has built the infrastructure layer that the most demanding AI workloads in the world depend on — GPU-dense clusters trusted by frontier labs, now a public company at the center of the AI compute era. That matters because the engineers building on top of that infrastructure need a developer platform that matches the performance bar of the underlying cloud. When I read the Developer Productivity PM role, I recognized the exact problem space I have spent the last several years working inside: instrumenting complex developer workflows, turning telemetry into investment decisions, and shipping the platform capabilities that let engineering organizations move faster without sacrificing quality. **Technical and AI/ML Foundation** My technical foundation is hands-on and spans both the infrastructure and AI layers that define this role. At Intuit, I owned the ICE platform — the internal developer platform serving engineers across QuickBooks, TurboTax, Mint, Mailchimp, and Credit Karma. I drove 275% YoY growth in ICE engagements, scaling to 675M+ in FY23, and led the throughput migration from 6K to 50K TPS via rSocket, supporting approximately 1.5M concurrent connections at sub-25ms TP99. I worked directly with SQL and BigQuery telemetry to surface developer pain points across ~20 mobile apps and 30+ product SKUs, and I built Asterias, a declarative asset lifecycle management platform with a GraphQL API, to give engineering leadership self-serve visibility into platform health. I also 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 over $1M in projected opex growth. On the AI evaluation side, I built aeval, a local-first model evaluation platform with five core eval types (factuality, reasoning, instruction-following, safety, code generation), adversarial safety testing with refusal detection, and statistical rigor including bootstrap confidence intervals, Welch's t-test, Cohen's d effect size, and saturation detection — with CI/CD integration and automated safety gates. The stack runs on FastAPI, TimescaleDB, Redis, and Ollama. I also built a 3-phase RL post-training 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 running real TRL-powered GRPO/DPO training with live SSE metric streaming on Apple Silicon (MPS) 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 — which means I understand CoreWeave's customer use cases from the inside, not just from a product spec. Earlier in my career at Splunk, I owned Search Service (Go microservices), Search Catalog (PostgreSQL metadata service), and SPL/SPL2 — building product roadmaps and prioritization frameworks for three microservice backlogs serving Fortune 500 customers. At Kaiser Permanente, I led the enterprise rollout of Splunk Logging-as-a-Service at 1.7 TB daily volume across 200+ internal customers and built Redis-backed caching infrastructure for scalability and fault tolerance. **Why This Role** The arc from owning a developer platform at Intuit's scale, to building AI evaluation infrastructure independently, to running RL training workloads on GPU clusters maps directly to what this role requires: someone who can instrument the full PDLC, design rigorous AI-impact evaluations, and translate that data into a credible roadmap for an Agentic Engineering Platform. CoreWeave's engineering organization operates at a scale and reliability bar that few companies match, and the opportunity to define how developer productivity is measured and improved in that environment — at the intersection of AI tooling and infrastructure — is exactly the kind of 0-to-1 platform problem I am built for. **Role-Specific Connection** The JD's emphasis on independently designing and executing AI evaluations that surface untapped opportunities across the PDLC is the part of this role I find most compelling — it is precisely the problem aeval was built to solve, applied now to an engineering organization rather than a model benchmark. The expectation that the PM will own dashboard design and socialize insights to senior leadership maps to my experience building Asterias and using BigQuery telemetry to drive CTO-level language investment decisions at Intuit. And the vision for an Agentic Engineering Platform aligns with the multi-agent orchestration work I have done with OpenClaw — designing gateway protocols, subagent delegation, and session management for coordinated AI workflows. **Selected Prior Experience** - Scaled ICE developer platform to 675M+ engagements in FY23 (275% YoY growth); drove throughput from 6K to 50K TPS via rSocket migration supporting ~1.5M concurrent connections at sub-25ms TP99. - Delivered ICE Self-Service (DevPortal, GitOps config, ICE Playground), reducing developer onboarding from 2–3 weeks to minutes in pre-prod and <24 hours for production; mitigated $1M+ in projected opex growth. - Used SQL and BigQuery telemetry to prioritize developer pain points across ~20 mobile apps and 30+ product SKUs; built Asterias, a declarative asset lifecycle management platform with GraphQL API. - Conducted enterprise-wide Service Language Assessment across 9 languages, analyzing usage data and developer feedback to inform strategic investment decisions presented to CTO. - Built aeval evaluation platform with bootstrap confidence intervals, Welch's t-test, Cohen's d effect size, saturation detection, and automated CI/CD safety gates — FastAPI, TimescaleDB, Redis, Ollama. - Built RL post-training workbench benchmarking TRL, VeRL, OpenRLHF, and NeMo RL across 12 algorithms with standardized throughput/memory/convergence metrics and GPU Docker passthrough. - Owned Search Service (Go microservices) and SPL/SPL2 at Splunk; designed RICE-based prioritization framework for three microservice backlogs serving Fortune 500 customers. - Led Splunk Logging-as-a-Service enterprise rollout at Kaiser Permanente (1.7 TB daily volume, 200+ internal customers); built Redis-backed caching for scalability and fault tolerance. **Closing** CoreWeave's mission — turning compute into capability for the pioneers building the next generation of AI — depends on an engineering organization that can ship at the speed the market demands. Developer productivity is not a support function at a company like CoreWeave; it is a strategic lever. I would bring to this team a combination of platform PM depth, hands-on data and AI evaluation experience, and direct familiarity with the GPU workloads your customers run — and I would be glad to discuss how that maps to the specific roadmap challenges your team is working through. Thank you for your consideration. --- **O. Felix Amoruwa** famoruwa@berkeley.edu | 909-731-9011 | felixamoruwa.info