DCIM Analyst - Infrastructure Operations

Cloudflare•San Francisco, CA

16h

About The Position

We are seeking a DCIM Analyst to be the data scientist for the Physical Layer, responsible for the integrity, forecasting, and visualization of all infrastructure datasets—Space, Power, Cooling, Cabling/Ports, and Asset Inventory. This technical role is part of the Infrastructure Operations organization, which is responsible for building, scaling, and running one of the world's largest and most important cloud networks. Cloudflare’s global network spans more than 330 cities, and is a key strategic asset that supports all of our customers and products. The DCIM Analyst is the architect of our physical intelligence. You own the complete analytical scope of the Nlyte platform—spanning Space, Power, Cooling, Connectivity, and Asset Lifecycle. You will move beyond simple monitoring to build an infrastructure health engine, serving as the mandatory "Validator" for all global changes. You will transform fragmented data into a unified capacity strategy, ensuring our edge network scales efficiently while safeguarding against resource exhaustion and physical risk. We operate in a fast-paced environment where you will be expected to drive both project delivery and operational excellence through continuous improvement, standardization, and optimization. This isn't just about day to day operations; it's about building a scalable, performant, secure, and resilient infrastructure that plays a critical role in us building a better Internet.

Requirements

Expert DCIM Analytics: 4+ years of experience administering the analytics module of a major DCIM platform (Nlyte, Sunbird, or similar). Must demonstrate the ability to build custom reports, not just use default dashboards.
Multi-Constraint Modeling: Proven experience modeling capacity across four distinct constraints: Space (Rack Units/Footprint), Power (kW draw vs. Circuit limits), Cooling (BTU/h and Airflow), Connectivity (Port density and Cabling availability).
Data Normalization: Experience managing data ingestion from varied hardware sources using standard protocols and normalizing that data for historical analysis.
BI Visualization: Proficiency in SQL and data visualization tools (e.g., Tableau, Grafana, PowerBI) to create the "Single Source of Truth" reporting for Finance and Strategy stakeholders
Domain Knowledge Deep understanding of the physical environment. You must understand why a rack is overheating, not just report that it is hot.
Power Distribution Architectures: Knowledge of data center power chains
Structured Cabling Standards: Familiarity with fiber/copper standards to accurately model port capacity and connectivity meshes.
Change Management Logic: Experience defining the business logic for Automated Capacity Validation—writing the rules that determine if a ticket is automatically approved or rejected based on data.
Root Cause Analysis: Experience using historical time-series data to perform forensic analysis after an incident (e.g., correlating a power drop with a specific server failure.
Principled: You have the confidence to act as a neutral arbiter. If the data shows a deployment is unsafe, you will withhold validation approval, regardless of pressure from deployment teams.
Curious: You proactively hunt for inefficiencies that others miss, treating the infrastructure as a puzzle to be optimized.

Responsibilities

Serve as the required approval step in the Change Management workflow. You must validate every proposed Move, Add, and Change (MAC) against real-time capacity constraints before the Administrator can issue a work order.
Enforce a "Zero-Overprovisioning" policy by blocking requests that breach redundancy thresholds for Space, Power, Cooling, or Network Port availability.
Develop forward-looking capacity models to forecast resource exhaustion.
Run "What-If" scenarios to determine the optimal placement of new high-density hardware (e.g., AI/GPU clusters) to avoid creating hot spots or stranded capacity.
Advise the DCIM Manager and Capacity Team on when and where to purchase additional colocation space or power based on consumption trends.
Design and own the data ingestion strategy for the Nlyte Real-Time Monitoring module. Ensure continuous polling of thousands of sensors across IT devices and facility equipment (CRACs, UPS, PDUs).
Manage the normalization of raw telemetry data from diverse protocols into a clean, actionable Time-Series Database.
Analyze the integrity of the Asset Management database. Identify "ghost servers" (powered on but not in inventory) and track asset aging to predict decommissioning waves.
Reconcile data discrepancies between "Discovered" network data and "Managed" inventory data, flagging errors for the Administrator to fix.
Transform raw data into executive-level dashboards. Calculate and report on critical efficiency metrics, including Power Usage Effectiveness (PUE) and carbon impact.
Define and tune global alerting thresholds to ensure operations teams are alerted to genuine risks without suffering from alert fatigue.