cover_letter / art_XoBJ-R9hssI

Dear Baseten Hiring Team, Baseten is doing something genuinely hard: making mission-critical AI inference reliable enough that companies like Cursor, Notion, and Abridge stake their products on it. That matters because the gap between a model that works in a notebook and one that serves millions of requests at sub-100ms latency is where most AI ambitions stall. My interest in this role is grounded in a specific experience: at Intuit, I inherited a developer platform that was assembled deal-by-deal — every new team got a bespoke onboarding path, bespoke configurations, bespoke exceptions. I turned that into a supported, priced product called ICE Self-Service, and the result was 275% YoY engagement growth scaling to 675M+ engagements in FY23. That pattern — converting one-off exceptions into sellable, supported capabilities — is exactly what Baseten is asking for in this Enterprise PM role. **Technical and Platform Foundation** My credibility in this space comes from building at both the infrastructure and product layers. At Intuit, I owned the platform that scaled inference-adjacent workloads from 6K to 50K TPS via an rSocket migration supporting approximately 1.5M concurrent connections with sub-25ms TP99 — the kind of throughput and latency profile that maps directly to what Baseten's inference customers require. I delivered the ICE Self-Service platform end-to-end: DevPortal, GitOps configuration, and ICE Playground, reducing developer onboarding from 2–3 weeks to minutes in pre-production and under 24 hours for production, while mitigating over $1M in projected opex growth. I also led a Mailchimp GCP-to-AWS migration for MSaaS, which required coordinating cross-cloud deployment models, updated DevPortal documentation, and production deadline accountability — the kind of multi-cloud deployment work that enterprise customers increasingly require from infrastructure vendors. On the security and access control side, I built the MSaaS Drift Detection and Resolution program: a Java JAR library that scanned Git repositories for configuration drift, paired with a DevPortal UI and a remediation roadmap using OpenRewrite. At Kaiser Permanente, I led the enterprise rollout of Splunk Logging-as-a-Service handling 1.7 TB of daily volume across 200+ internal enterprise customers, and built caching infrastructure using Redis and XC10 to address scalability, fault tolerance, and data redundancy — the operational concerns that enterprise IT and security teams scrutinize in procurement. Beyond platform infrastructure, I have hands-on depth in the ML layer that is increasingly non-negotiable for credibility in this space. I built a production RL post-training workbench covering GRPO, DPO, PPO, and nine other algorithms across TRL, VeRL, OpenRLHF, and NeMo RL — with live SSE metric streaming, GPU Docker passthrough, and standardized throughput/memory/convergence benchmarking. I built aeval, a local-first model evaluation platform with FastAPI orchestration, TimescaleDB, Redis job queuing, and automated safety gates with CI/CD regression detection. These are not survey-level projects; they are the kind of hands-on builds that let me have a real conversation with the systems engineers at Baseten about what enterprise customers are actually asking for when they request compliance posture or deployment flexibility. **Why This Role** Baseten's April 2026 RBAC rollout and the Frontier Gateway launch signal that the enterprise surface is being built now — and the job description is honest that today it is assembled deal-by-deal. That is the exact problem I have solved before, and I want to solve it again at a company operating at the frontier of inference infrastructure. What excites me specifically is the scope: deployment models, compliance posture, identity and access, and billing and spend controls treated as a unified product surface rather than a checklist. Baseten's customer base — regulated industries like healthcare (Abridge) and financial services — brings exactly the procurement and security requirements I have navigated at Kaiser Permanente and Intuit. The opportunity to define what a complete Baseten Enterprise offering looks like, working directly with founders and infrastructure engineers, is the kind of 0-to-1 product work I do best. **Selected Relevant Experience** - 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, mitigating $1M+ in projected opex growth — converting bespoke onboarding into a supported, priced capability. - Scaled ICE platform throughput from 6K to 50K TPS via rSocket migration supporting ~1.5M concurrent connections with sub-25ms TP99; achieved 275% YoY growth scaling to 675M+ engagements in FY23. - Led Mailchimp GCP-to-AWS migration for MSaaS, delivering Golang template, MySQL persistence integration, and updated DevPortal documentation to meet production deadline — multi-cloud deployment execution under enterprise constraints. - Initiated MSaaS Drift Detection and Resolution program: wrote Java JAR library to scan Git repos for configuration drift, partnered with Design on DevPortal UI, and built remediation roadmap using OpenRewrite — security and compliance as product, not checkbox. - Led enterprise rollout of Splunk Logging-as-a-Service (1.7 TB daily volume, 200+ internal enterprise customers) and ITSI Application Monitoring-as-a-Service at Kaiser Permanente. - Built RL post-training workbench benchmarking GRPO/DPO/PPO across TRL, VeRL, OpenRLHF, and NeMo RL with GPU Docker passthrough and live metric streaming — hands-on ML infrastructure depth. - Architected multi-provider LLM orchestration (Claude, GPT-4, Gemini) with fallback routing, structured output validation, and token budget optimization at Fintellect AI — production model serving tradeoffs in practice. - Conducted enterprise-wide Service Language Assessment across 9 languages, analyzing usage data and developer feedback to inform strategic investment decisions presented to Intuit CTO — the kind of rigorous, data-driven research that drives roadmap calls. **Closing** Baseten's mission — making frontier AI inference reliable enough for the world's most demanding companies — only succeeds if the enterprise layer is a product, not a negotiation. I have spent twelve years building exactly that kind of platform infrastructure, and I have the ML depth to be credible with the engineers who build it and the enterprise buyers who require it. 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