cover_letter / art_-VToNFRKYxQ

Dear Fivetran / dbt Labs Hiring Team, Fivetran and dbt Labs are building the infrastructure layer that makes data trustworthy from the moment it moves through every transformation — the kind of foundational work that quietly powers how thousands of organizations make decisions and train AI systems. That mission resonates with me directly: I spent three years at Intuit as a Staff PM owning developer frameworks and platform infrastructure, watching firsthand how the quality of developer tooling determines whether engineers ship with confidence or spend their days fighting friction. That experience — combined with the AI and ML work I've done since — is exactly why this role caught my attention. **Technical and AI Foundation** My technical foundation spans both the engineering and product sides of developer platforms. At Intuit, I owned the ICE platform — the internal compute and event infrastructure serving QuickBooks, TurboTax, Mint, Mailchimp, and Credit Karma. I extended Java and Python SDK Starter Kits with scaffolding templates, build configurations (Gradle/Maven), testing frameworks, and CI/CD integration so developers could go from zero to production-ready microservice in minutes. I 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. That platform scaled to 675M+ engagements in FY23, and I drove a rSocket migration that took throughput from 6K to 50K TPS supporting approximately 1.5M concurrent connections at sub-25ms TP99. These weren't abstract roadmap exercises — they required working fluently with engineering partners on system architecture, API contracts, and the boundary between platform capability and product surface. On the AI side, I've been building and studying AI-assisted development workflows hands-on. I built an RL post-training workbench that benchmarks GRPO and DPO across TRL, VeRL, OpenRLHF, and NeMo RL — implementing 12 RL algorithms (PPO, GRPO, DAPO, REINFORCE, DPO, SimPO, and others) with live SSE metric streaming, GPU Docker passthrough, and cross-framework throughput/memory/convergence benchmarking. I built the OpenClaw multi-agent orchestration framework with a gateway protocol, subagent delegation, profile management, and session switching — coordinating AI agent workflows across multiple domains. I've also built aeval, a local-first model evaluation platform with adversarial safety testing, bootstrap confidence intervals, Welch's t-test, and CI/CD regression detection. These projects gave me a practitioner's understanding of what agentic UX actually requires: context management, structured output validation, fallback routing, and the workflow design decisions that determine whether an AI assistant accelerates or interrupts a developer. My NeurIPS 2014 publication on neural networks for protein secondary structure prediction — and the 2026 rewrite of that system from a hand-coded C++ BPTT implementation to a PyTorch platform spanning 413 parameters to 8B — reflects a long arc of engagement with ML that goes well beyond product-layer familiarity. **Why This Role** The dbt Cloud developer experience sits at exactly the intersection I've been working toward: a developer-facing product where AI isn't a bolt-on feature but the primary mechanism for lowering the floor for new practitioners while raising the ceiling for experts. The Developer Agent, Studio IDE, Cloud CLI, and VS Code Extension together represent a coherent developer workflow — and the tight feedback loop with the Fusion team to surface SQL comprehension, compute, and context through developer-facing UX is the kind of platform-product boundary work I find most interesting. What excites me specifically about this role is the challenge of defining what a best-in-class AI coding assistant looks like for data practitioners — a population that spans seasoned analytics engineers writing complex SQL transformations to newer practitioners who need guided, context-aware development experiences. I've thought carefully about what Cursor and Copilot get right (low-latency inline suggestions, strong context retrieval, minimal workflow interruption) and where they fall short for domain-specific workflows. Translating those instincts to the dbt Cloud context — where the user's mental model is SQL-first, transformation-centric, and deeply tied to warehouse semantics — is a genuinely interesting design problem. I'm also drawn to the Cloud CLI work: making a CLI a first-class product that developers actively choose requires the same discipline as IDE work, and I have direct experience building developer tooling that changed adoption curves. **Selected Relevant Experience** - **ICE Self-Service Platform (Intuit):** Delivered DevPortal, GitOps config, and ICE Playground — reducing developer onboarding from 2–3 weeks to minutes in pre-prod and under 24 hours for production; mitigated $1M+ in projected opex growth. - **SDK Starter Kits (Intuit):** Extended Java and Python SDKs with scaffolding templates, build configurations (Gradle/Maven), testing frameworks, and CI/CD integration — enabling developers to reach production-ready microservices in minutes. - **Platform Scale (Intuit):** Achieved 275% YoY growth in ICE engagements, scaling to 675M+ in FY23; drove rSocket migration from 6K to 50K TPS supporting ~1.5M concurrent connections at sub-25ms TP99. - **OpenClaw Multi-Agent Orchestration (StreamIO AI):** Built multi-agent framework with gateway protocol, subagent delegation, profile management, and session switching — enabling coordinated AI agent workflows; directly relevant to Developer Agent agentic workflow design. - **RL Post-Training Workbench:** Built 3-phase workbench covering the full RLHF/DPO pipeline with live SSE metric streaming, 12 RL algorithm implementations, and head-to-head framework benchmarking (TRL, VeRL, OpenRLHF, NeMo RL) — establishing hands-on fluency with the AI systems underlying coding assistants. - **aeval — AI Model Evaluation Platform:** Built local-first evaluation platform with adversarial safety testing, statistical rigor (bootstrap CIs, Welch's t-test, Cohen's d), and CI/CD regression detection — directly applicable to measuring Developer Agent quality and safety. - **Search Orchestration (Splunk):** Owned Search Service (Go microservices), Search Catalog (PostgreSQL), and SPL/SPL2 — delivered Scheduler Service end-to-end in ~4 months; achieved up to 10x query performance improvements for beta enterprise customers. **Closing** Fivetran and dbt Labs are building the data infrastructure layer that AI systems and analytics teams will depend on for the next decade. The developer experience you ship today — how quickly a practitioner can write, debug, and trust a dbt model — directly determines how broadly that infrastructure gets adopted and how much value it creates. I'd bring to this role 12+ years of developer platform and AI product experience, a hands-on technical foundation that lets me work as a genuine partner to engineering, and a specific point of view on what AI-assisted development should feel like for domain-specific practitioners. I'd welcome the conversation. Sincerely, **O. Felix Amoruwa** famoruwa@berkeley.edu | 909-731-9011 | felixamoruwa.info