AI / Cloud FinOps Lead (COGS Visibility & Optimization)

Rox Data Corp•San Francisco, CA

About The Position

Rox is building the AI-native revenue operating system for modern go-to-market teams. Backed by Sequoia, GV, and General Catalyst, we’re working with some of the most ambitious enterprise teams to replace fragmented CRM workflows with intelligent, autonomous systems. Rox connects data across the GTM stack, deploys AI agents to do real work, and gives revenue leaders a clear, shared picture of what actually drives outcomes. We’re a small, fast-moving Series A team taking on one of software’s most entrenched categories — and we’re winning by combining deep technical rigor with obsessive focus on usefulness. About the Finance Team Finance at Rox is not a back-office function — it’s an operating partner to Engineering, Product, and Go-To-Market. The team builds the systems that make performance measurable, tradeoffs explicit, and scaling decisions repeatable. As Rox scales AI-driven workloads, COGS visibility and optimization become foundational to how we manage gross margin, pricing, and product strategy. This role sits at the center of that effort. About the Role This is a founding-level AI / Cloud FinOps Lead role with end-to-end ownership of COGS visibility and optimization. You will sit at the intersection of Finance, Engineering, and Product, responsible for making costs measurable, attributable, and actionable — without compromising latency, reliability, or customer experience. This role is both deeply analytical and highly operational: you’ll build the cost model, establish trusted metrics, and translate insight into concrete optimization initiatives that leadership acts on. If you enjoy turning complex AI and cloud cost systems into clear decision frameworks — and want real ownership — this role is built for that.

Requirements

FinOps & Cloud Cost Expertise
Experience owning cloud cost management, COGS attribution, or unit economics in cloud-heavy (ideally AI-heavy) environments.
AI / ML Cost Literacy
Strong understanding of AI/ML cost dynamics — including token economics, model inference, and RAG-related costs — and how to translate them into actionable unit economics.
Data Fluency
High proficiency in SQL and comfort working directly with raw billing exports and warehouse tables.
BI & Metrics Ownership
Ability to build trusted, well-defined metrics and views in modern BI tools (Sigma, Looker, Tableau, etc.).
Strategic Communication
Ability to translate technical cost data into clear financial narratives and decision frameworks for leadership.

Nice To Haves

Familiarity with infrastructure-as-code, observability tooling, or engineering telemetry systems
Experience with major cloud platforms (AWS, GCP, Azure)
Background in FP&A or strategic finance with a strong technical bias

Responsibilities

Define the Cost Model
Partner with Engineering to specify instrumentation, pipelines, and attribution so COGS can be tracked at the right granularity — by customer, workflow/action, request, model, token type, and external dependencies (e.g., AI search APIs), across compute, storage, and network.
Build the Analytics Layer
Own the source of truth for unit economics in our BI stack (Sigma), with trusted definitions for usage and cost metrics such as tokens per action, $/token, $/request, and $/customer or workflow.
Profitability & Decision Support
Build profitability views by customer, model, and workload to support pricing, packaging, and product decisions using real unit economics.
Drive Optimization Strategy
Translate data into concrete initiatives — model routing, prompt and context limits, caching strategies, infrastructure tuning, and vendor usage changes — with estimated savings and explicit cost vs. latency or quality tradeoffs.
Orchestrate Follow-Through
Establish a repeatable weekly and monthly cadence for unit-economics and profitability reviews; provide leadership with clear decision support on major tradeoffs.
Anomaly Detection & Deep Dives
Create variance detection to surface cost spikes and regressions early; run deep dives and validate findings closely with Engineering.