← nvidia / Product Manager, AI Platform SW - Agentic AI Kernel Generation
tailored_resume / art_IPYr81CNdRA
role
model
claude-sonnet-4-6
created
2026-05-18T23:26
Content
# O. FELIX AMORUWA famoruwa@berkeley.edu | 909-731-9011 | felixamoruwa.info Technical Product Leader with 12+ years building developer-facing platforms and AI frameworks at scale — from hand-coding BPTT in C++ (2004) to building RL post-training workbenches that benchmark GRPO/DPO across TRL, VeRL, OpenRLHF, and NeMo RL today. NeurIPS published researcher. Deep experience shipping agentic AI products, automated evaluation suites, and developer platform infrastructure at scale (675M+ engagements at Intuit). --- ## PROFESSIONAL EXPERIENCE ### STREAMIO AI | Oakland, CA **Founder & CEO — streamio.ai (macOS, Linux, iOS & Web)** 09/24 – Present - Implemented OpenClaw multi-agent orchestration framework with gateway protocol, subagent delegation, profile management, and session switching — enabling coordinated AI agent workflows across real estate, insurance, health/dental, and financial markets industries. - Shipped MCP server exposing screen capture tools to AI coding assistants, with embedded terminal (xterm.js/node-pty, 20+ slash commands) and native macOS ScreenCaptureKit integration via Swift — connecting agents directly to execution environments. - Built production Electron + React + TypeScript desktop application with 100+ components, Redux Toolkit state management, and native macOS/Linux support — providing real-time AI analysis of screen captures using Claude (MCP SDK). - Led 0-to-1 product strategy, AI development, and go-to-market execution including customer discovery interviews and iterative product refinement. ### FINTELLECT AI | Oakland, CA **Founder & CEO — fintellectai.io (iOS/Android/Web)** 09/24 – Present - Built domain-specific Fintellect Agents — conversational AI agents embedded in the mobile and web app, each scoped to distinct financial focal points, delivering guided, context-aware advisory interactions. - Architected RAG retrieval pipeline with ChromaDB vector store, multi-provider LLM orchestration (Claude, GPT-4, Gemini) with fallback routing, structured output validation, and token budget optimization. - Built AI-powered charting and macroeconomic analysis tools to optimize asset class selection, execution planning, and risk management; integrated LLM models for automated trade analysis at scale. ### INTUIT | Mountain View, CA **Staff Product Manager — Developer Frameworks & Platform Infrastructure** 06/21 – 09/24 - 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. - Delivered ICE Self-Service platform (DevPortal, GitOps config, ICE Playground), reducing developer onboarding from 2–3 weeks to minutes in pre-prod and <24 hours for production, while mitigating $1M+ in projected opex growth. - Extended Java and Python SDK Starter Kits with scaffolding templates, build configurations (Gradle/Maven), testing frameworks, and CI/CD integration — empowering developers to go from zero to production-ready microservice in minutes. - Conducted enterprise-wide Service Language Assessment across 9 languages (Java, Python, Kotlin, Go, TypeScript, Scala, PHP, C++, Groovy), analyzing usage data and developer feedback to inform strategic language investment decisions presented to CTO. - Worked closely with telemetry and usage data (SQL, BigQuery) to prioritize developer pain points across ~20 mobile apps and 30+ product SKUs; built Asterias, a declarative asset lifecycle management platform with GraphQL API. - 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. - Implemented ICE Presence in async chat, generating $480K/month in additional invoicing; deployed Background-to-Foreground Messaging on iOS/Android with <100ms latency. DeveloperWeek 2022 speaker. ### SPLUNK INC. | San Francisco, CA **Senior Product Manager — Search Orchestration** 01/19 – 01/21 - Led query performance optimization initiative for beta customer (Assurance), building mirrored Enterprise topology for benchmark testing and achieving up to 10x performance improvements in SCS search. - Owned Search Service (Go microservices), Search Catalog (PostgreSQL metadata service), and Splunk Processing Language (SPL/SPL2) — built product roadmaps, PRDs, user stories, and acceptance criteria for Splunk Cloud Services (SCS). - Delivered Scheduler Service end-to-end in ~4 months (June–Oct 2019), enabling scheduled search capabilities for first-party applications; demoed at Splunk .conf19. - Designed repeatable RICE-based prioritization framework for 3 microservice backlogs, balancing internal partner, third-party developer, and Fortune 500 customer requirements. Splunk .conf18 Speaker. ### KAISER PERMANENTE | Pleasanton, CA **SOA Technical Product Manager** 08/12 – 12/18 - Led development and enterprise rollout of Splunk Logging-as-a-Service (1.7 TB daily volume, 200+ internal enterprise customers) and ITSI Application Monitoring-as-a-Service. - Built caching capability using Redis and XC10 across the enterprise to address scalability, fault tolerance, and data redundancy challenges. - Led demand forecasting and capacity planning for SOA product suite, analyzing demand drivers and developing IT capacity plans across multiple datacenters. ### IBM | Pittsburgh, PA / Foster City, CA / RTP, NC **Software Engineer — Business Intelligence Products** 07/05 – 03/10 - Resolved multiple $1–5M high-severity customer escalations for enterprise BI software products; led cross-functional root cause analysis and customer-facing resolution, improving case resolution time by 20% across the Support Organization. --- ## AI/ML RESEARCH & PROJECTS ### RL Workbench — Post-Training RL Platform (2026) - Built 3-phase RL post-training workbench covering the full RLHF/DPO pipeline: Reward Lab for designing and A/B testing reward functions (RLVR, learned, hybrid) across 4 datasets (GSM8K, MATH, HumanEval, UltraFeedback), Playground for real TRL-powered GRPO/DPO training with live SSE metric streaming on Apple Silicon (MPS) or CUDA, and Arena for head-to-head framework benchmarking (TRL, VeRL, OpenRLHF, NeMo RL) with GPU passthrough in Docker containers. - Implemented 12 RL algorithms (PPO, GRPO, DAPO, REINFORCE, REINFORCE++, RLOO, DPO, SimPO, IPO, KTO, ORPO, SPPO) with algorithm-specific metric profiles, cross-tab workflow lineage tracking, and standardized throughput/memory/convergence benchmarking across frameworks. ### aeval — AI Model Evaluation Platform (2025–2026) - Built local-first evaluation platform with 5 core eval types (factuality, reasoning, instruction-following, safety, code generation), adversarial safety testing with refusal detection, and data contamination detection via SHA-256 hashing. - Statistical rigor: bootstrap confidence intervals, Welch's t-test, Cohen's d effect size, saturation detection. CI/CD integration with regression detection and automated safety gates. Stack: FastAPI orchestrator, TimescaleDB, Redis job queue, Next.js dashboard, Ollama. ### AutoEval — Automated Visual Evaluation for Robot Model Training (2025) - Built automated visual evaluation system for robot model training by repurposing StreamIO's screen capture + HLS streaming + multimodal AI pipeline to score model outputs (grasp poses, segmentation maps, bounding boxes) against natural-language evaluation rubrics — reducing evaluation cycles from 72 hours to ~4 minutes. - Zero-integration architecture captures from any visualization tool (RViz, Matplotlib, custom dashboards) via screen capture; multimodal AI (Claude/GPT-4V) performs spatial reasoning on prediction frames, generating structured PASS/FAIL reports with confidence scores and edge-case flagging. ### BRAIN — Protein Structure Prediction ML Platform (UC Berkeley 2004; rewritten 2026) - Built production ML platform in PyTorch with 5 neural architectures (feedforward, GRU, Transformer, ESM-2, multi-task), MLflow experiment tracking, Optuna HPO, and FastAPI serving — 823 automated tests, Docker orchestration (6 containers). - Original 2004 system: hand-coded neural network in C++ with custom BPTT; 2026 rewrite spans 413 to 8B parameters (19M-fold scale increase). ### NeurIPS 2014 — Montreal, Canada - Accepted paper on artificial neural networks for protein secondary structure prediction. ### Lawrence Berkeley National Laboratory - Computational analysis of RNA/DNA sequences under Dr. Steven Holbrook; assessed secondary structure accuracy using Monte Carlo and AI algorithms. --- ## EDUCATION **CARNEGIE MELLON UNIVERSITY, SILICON VALLEY** Master of Science — Software Management Mountain View, CA (08/16) **CARNEGIE MELLON, TEPPER SCHOOL OF BUSINESS** Master of Business Administration — Concentrations: Finance, Quantitative Analysis, Economics Pittsburgh, PA (03/12) **UC BERKELEY, COLLEGE OF ENGINEERING & HAAS SCHOOL OF BUSINESS** Bachelor of Science — Computational Engineering Science Berkeley, CA (12/05) Certificate — Corporate Finance for Financial Engineering --- ## TEACHING EXPERIENCE ### DE ANZA COLLEGE | Cupertino, CA **Adjunct Faculty — Computer Information Systems Department** 04/18 – Present - CIS 35A/B, 36A/B — Java Programming; CIS 64C/64G — Introduction to PL/SQL; CIS 44F — Introduction to Data Analytics; CIS 64E — Fundamentals of Large Scale Cloud Computing (AWS, GCP); CIS 102 — Ethical Hacking; CIS 95F — Managing Cloud Projects. --- ## ADDITIONAL INFORMATION - O'Reilly & OER Publications: Data Analytics and Managing Cloud Projects (forthcoming Summer 2026) - FAA Commercial Drone License; FAA Private Pilot License (intended CFI) - First Tee Assistant Golf Coach — San Francisco Chapter (2012–Present), Pittsburgh Chapter (2011–2012) - Interests: Classical & Jazz Saxophone (Alto/Tenor/Baritone), Triathlon, Golf, Intramural Basketball & Soccer