cover_letter / art_rlG8mPuMFhg

Dear Okta Enterprise AI Hiring Team, Okta sits at the precise intersection where identity infrastructure and AI converge — and that intersection is where the next decade of enterprise security will be decided. The mission of securing every identity, from AI agents to humans, is not abstract to me: I have spent the last two years building multi-agent orchestration systems, RAG pipelines, and AI evaluation frameworks from scratch, and I understand firsthand how quickly agentic systems can outpace the trust and governance structures meant to contain them. That gap is exactly what Okta is positioned to close. **Technical and AI Foundation** My AI work is hands-on and spans the full stack. In 2025–2026, I built **aeval**, a local-first AI model evaluation platform with five core eval types — factuality, reasoning, instruction-following, safety, and code generation — backed by adversarial safety testing with refusal detection, bootstrap confidence intervals, Welch's t-test, Cohen's d effect size, and automated safety gates integrated into CI/CD pipelines. This is precisely the kind of AI quality measurement and evaluation framework Okta's JD calls out. The stack runs on FastAPI, TimescaleDB, Redis, and Ollama, with a Next.js dashboard for real-time visibility. I also built an **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 datasets; a Playground for live TRL-powered GRPO/DPO training with 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 containers. I implemented 12 RL algorithms with standardized throughput, memory, and convergence benchmarking. This is not academic familiarity — it is production-grade tooling I designed, built, and operate. On the agentic side, I built **OpenClaw**, a multi-agent orchestration framework with a gateway protocol, subagent delegation, profile management, and session switching — enabling coordinated AI agent workflows across real estate, insurance, health/dental, and financial markets verticals. I also architected a RAG retrieval pipeline at Fintellect AI with ChromaDB vector store, multi-provider LLM orchestration across Claude, GPT-4, and Gemini with fallback routing, structured output validation, and token budget optimization. These are the agentic system design patterns Okta's JD lists as a nice-to-have; for me they are table stakes. This technical depth is grounded in a research foundation: my work on neural networks for protein structure prediction was accepted at **NeurIPS 2014**, and my original hand-coded BPTT system in C++ from UC Berkeley in 2004 was rewritten in 2026 as a production PyTorch platform spanning 413 parameters to 8B — a 19-million-fold scale increase across five architectures with MLflow, Optuna HPO, and FastAPI serving. **Why This Role** My career arc — from software engineering and ML research to platform product management at scale to founding AI-native companies — maps directly onto what Okta's Enterprise AI Engineering Team is building: products that start as 0→1 innovations and must be scaled into reliable, company-wide capabilities with enterprise rigor. I have lived both sides of that transition. What specifically draws me to this role is the Customer Zero model. Okta's TDI team is not building AI features in isolation — it is designing and shipping AI-powered products that simultaneously reshape internal operations and inform what Okta brings to market. That feedback loop between internal deployment and external product strategy is the most accelerated learning environment in enterprise AI, and it is where I want to operate. The challenge of defining quality metrics for AI agents, evaluating performance across copilots and workflow automation, and partnering across IT, HR, Finance, and GTM to drive real adoption — that is a complete product problem, not a feature problem. **Selected Relevant Experience** - **Intuit (Staff PM, Developer Frameworks & Platform Infrastructure):** 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. Scaled ICE engagements 275% YoY to 675M+ in FY23 across QuickBooks, TurboTax, Mint, Mailchimp, and Credit Karma. - **Intuit:** Scaled platform throughput from 6K to 50K TPS via rSocket migration supporting approximately 1.5M concurrent connections with sub-25ms TP99 — directly relevant to Okta's need for secure, scalable AI infrastructure at enterprise scale. - **aeval (AI Evaluation Platform):** Built production AI evaluation platform with adversarial safety testing, statistical rigor (bootstrap CI, effect size), and automated safety gates — directly addressing Okta's requirement to define quality metrics and evaluate AI performance. - **OpenClaw / Fintellect AI (Agentic Systems):** Designed and shipped multi-agent orchestration with gateway protocol and subagent delegation; built RAG pipeline with multi-provider LLM orchestration, fallback routing, and token budget optimization. - **Splunk (Senior PM, Search Orchestration):** Owned Go microservices, PostgreSQL metadata service, and SPL/SPL2; delivered Scheduler Service end-to-end in approximately four months; achieved up to 10x query performance improvements for Fortune 500 beta customers. - **StreamIO AI (Founder & CEO):** Led 0-to-1 product strategy, AI development, and go-to-market execution for a production Electron + React + TypeScript desktop application with real-time AI analysis, multi-agent workflows, and cross-platform deployment — demonstrating the full idea-to-adoption lifecycle the JD requires. - **Kaiser Permanente (SOA Technical PM):** Led enterprise rollout of Splunk Logging-as-a-Service at 1.7 TB daily volume across 200+ internal enterprise customers — relevant experience operating as an internal Customer Zero deploying infrastructure at enterprise scale. **Closing** Okta's mission — securing every identity as AI proliferates — is the right problem to be working on, and the Enterprise AI Engineering Team is the right place to work on it. I bring 12+ years of platform product leadership, hands-on AI/ML engineering depth, and a track record of taking products from zero to scale in both enterprise and startup environments. I would welcome the opportunity to discuss how my background maps to what you are building. Thank you for your consideration. Sincerely, **O. Felix Amoruwa** famoruwa@berkeley.edu | 909-731-9011 | felixamoruwa.info