Principal Engineer - Data Platform

Company Description It all started when engineer Fred Luddy wrote code that automated a tedious task for his coworker, Phyllis. She cried tears of joy. That moment inspired Fred to build a company that could do that for everyone—freeing people from busywork so they could focus on meaningful work. Today, ServiceNow is the AI control tower for business reinvention. Our ServiceNow AI platform brings together any AI, any data, and any workflow— helping 85% of the Fortune 500® work smarter, faster, and better. We're building an AI-native culture where technology and talent are unstoppable together. And we're just getting started. Join us to put AI to work for people. Job Description Employees can work remotely Team Join the Global Cloud Services (GCS) organization's platform engineering team, which is building ServiceNow's next-generation data foundation on the modern stack: Trino for distributed queries, dbt for transformations, Iceberg for lakehouse architecture, and Argo Workflows for orchestration. At the center of that effort is the GCS Data Warehouse, the modern lakehouse that will replace the organization's existing Cloudera-based data platform and serve as the substrate every GCS data consumer is built upon. Role As Principal Engineer for Data Platform Modernization, you will be the foundational architect for the GCS Data Warehouse and everything upstream of the query layer: how data lands, how it is transformed, and how it is served as correct, well-governed tables on the lakehouse. You will own the target architecture and lead the program that moves ServiceNow's Global Cloud Services data off Cloudera (Impala, Hive, HDFS, Hive Metastore) onto the modern lakehouse (Trino, Iceberg, dbt, a modern catalog). The estate is real: hundreds of tables and pipelines feeding dozens of downstream consumers, with years of accumulated history, spread today across on-premises, virtualized, and cloud clusters in multiple regions. Your job is to design the new foundation, stand up the ingestion and transformation layers that feed it, and move those workloads onto it with verified correctness and zero data loss at petabyte scale. The platform runs in the public cloud, and it must operate across both commercial and regulated environments. ServiceNow's regulated footprint includes government and other controlled enclaves with their own isolation, data-residency, and compliance requirements, so this is not a single-account, single-region system. You will design one portable architecture that deploys independently into each boundary, commercial and regulated alike, carrying the right isolation, access control, and audit posture into each. Getting that portability right is one of the hardest and most defining parts of the role. This is a one-year program with a clear end state: Cloudera decommissioned, so the organization does not renew it. There is a concrete cost and consolidation mandate behind that deadline, and it shapes every decision. You will make the high-leverage architectural calls fast, sequence the work so the platform is proven incrementally rather than in one high-risk cutover, and keep it moving at pace. This is a hands-on technical leadership role, not a management role. You will define the architecture, set the correctness and quality bar, make the hardest technical decisions, and keep the platform coherent as it scales. You will not manage people; you will lead through architecture, deep technical judgment, and influence, partnering closely with the engineers building alongside you and with Engineering and GCS leadership. This is a unique opportunity to define the data foundation for all of Global Cloud Services at ServiceNow's scale, and to do it at startup velocity within a Fortune 500 environment. What you get to do in this role Design the GCS Data Warehouse, the modern lakehouse foundation (Trino, Iceberg, dbt, a modern catalog) that replaces the existing Cloudera-based platform and serves as the substrate for GCS data consumers. Lead the one-year program to move GCS data off Cloudera (Impala, Hive, HDFS, Hive Metastore) so the organization can decommission it rather than renew, sequencing the work as a phased, low-risk path with each workload verified on the new foundation before the old one is retired. Design one portable architecture that deploys independently into commercial and regulated environments, with the isolation, data-residency, access-control, and audit posture each boundary requires. Treat operating across those boundaries as first-class architecture, not a later hardening step. Own the ingestion architecture: change data capture from the primary source systems, transactional PostgreSQL databases, landed into Iceberg. This means log-based CDC off the Postgres write-ahead log, handling upstream schema evolution, at-least-once delivery and deduplication, late and out-of-order data, and the reconciliation of streaming changes with backfills into correct, queryable Iceberg tables (merge-on-read, compaction). Own the streaming layer that carries those changes. Kafka is already in the estate and is the incumbent; you will assess it and decide whether to carry it forward or replace it, weighing operational weight, ecosystem fit, portability across environments, and the one-year timeline. Define the data and schema translation approach: Hive and Impala schemas and partitioning onto Iceberg tables, legacy file formats onto the lakehouse, and HiveQL, Impala SQL, and Spark transformations onto Trino SQL and dbt models. Set the correctness bar: reconcile new outputs against the source platform as ground truth, with fail-loud validation so any divergence is caught before cutover, never discovered after. Petabyte-scale with zero data loss. Design data governance and security on the lakehouse: access control, sensitive-data handling, and audit on Iceberg and Trino, including how that posture differs across commercial and regulated boundaries and how it replaces the legacy Hive and Ranger model. This is a first-class design workstream, not a footnote. Help design the platform's operational model: the SLOs, observability, runbooks, and on-call approach that will keep it reliable in production once workloads are live. Establish engineering standards for reliability, determinism, observability, and production readiness, and hold the bar as workloads move onto the new foundation. Lead through influence: align the engineers building alongside you to the target architecture, review their designs, and resolve the hard technical tensions, without taking the keyboard away from them. Navigate enterprise constraints, security, compliance, and approval processes, while keeping the program moving at pace. Drive the responsible use of AI and ML tooling to accelerate migration, translation, and validation work. What You Get To Do In This Role Own the end-to-end technical architecture of the FinOps Engineering Platform, ensuring the GCS Data Warehouse, data platform, development platform, infrastructure, Forecast Engine, and FCR automation compose into one coherent, scalable system. Lead the design and development of the GCS Data Warehouse and the program to migrate ServiceNow's Global Cloud Services data platform off Cloudera onto the modern lakehouse, with zero data loss and verified correctness. Set the technical vision and multi-year roadmap for the platform, and translate it into the concrete standards and interfaces each workstream builds against. Make the highest-leverage, hardest-to-reverse technical decisions: technology selection, system boundaries, data contracts, and the architectural patterns that span workstreams. Establish platform-wide engineering standards for reliability, determinism, observability, security, and production readiness, and hold the bar across teams. Lead through influence: partner with the Senior Staff engineers who own each workstream, review their designs, resolve cross-team architectural tensions, and align everyone to a single technical direction. Drive innovation across the platform, including the responsible use of AI/ML tooling to accelerate development and improve platform capabilities. Foster a culture of engineering craftsmanship, knowledge-sharing, and thoughtful quality practices across every team building on the platform. Move fast: keep the platform shipping in tight, high-velocity loops while protecting the architectural integrity that lets it scale. Technical Leadership & Architecture Define the reference architecture for the FinOps Engineering Platform and the contracts between its parts: how the data platform serves the Forecast Engine, how forecasts drive FCR automation, how the development platform productionizes analytics, and how all of it runs on the shared infrastructure. Lead technical decision-making on the platform-wide technology stack, system boundaries, and architectural patterns, arbitrating trade-offs that no single workstream can resolve alone. Establish best practices for data modeling, simulation and forecasting, pipeline development, orchestration, and platform scalability across the modern data stack. Own the cross-cutting non-functional requirements: reliability, determinism and reproducibility, observability, security and compliance, performance, and cost. Drive innovation in FinOps data analytics and forecasting, evaluating and adopting emerging technologies where they raise the platform's ceiling. GCS Data Warehouse: Modernization & Cloudera Migration Lead the design of the GCS Data Warehouse, the modern lakehouse foundation (Trino, Iceberg, dbt, a modern catalog) that replaces the existing Cloudera-based platform (Impala, Hive, HDFS, Hive Metastore) and serves as the substrate for the entire FinOps Engineering Platform. Own the migration strategy and sequencing: a phased, low-risk path that moves workloads off Cloudera incrementally rather than in a single high-risk cutover, with the legacy platform decommissioned only once each workload is verified on the new foundation. Establish full inventory and lineage of the existing platform first, the tables, transformations, scheduled jobs, and downstream consumers (Tableau, Lightdash, pipelines, the Forecast Engine), so nothing is migrated blind and nothing is left stranded. Define the data and schema translation approach: Hive/Impala schemas and partitioning onto Iceberg tables, legacy file formats onto the lakehouse, and HiveQL/Impala SQL and Spark transformations onto Trino SQL and dbt models. Set the correctness bar for the migration: dual-run old and new in parallel and reconcile outputs against the source platform as ground truth, with fail-loud validation so any divergence is caught before cutover, never discovered after. Petabyte-scale with zero data loss. Plan and execute consumer cutover and the retirement of the Cloudera cluster, capturing the infrastructure cost savings (a FinOps win the platform itself can measure) and the operational simplification of consolidating onto one modern stack. Navigate enterprise constraints, security, compliance, and approval processes, while keeping the migration moving at pace. Platform Architecture Across Workstreams GCS Data Warehouse: The foundational lakehouse the whole platform sits on, and the migration that retires the legacy Cloudera platform onto it (see above). Analytics & cost-governance data platform: Guide the lakehouse architecture (Trino, dbt, Iceberg, Lightdash), data modeling for cost allocation and showback, query performance at scale, and metadata, lineage, and governance. Cloud development platform: Guide the notebook-to-production pathways (workspace provisioning, parameterization, validation, automated deployment) so exploratory analysis reaches production safely and quickly. Multi-cloud infrastructure, DevOps, and SRE: Guide the Kubernetes, IaC, CI/CD, security, and observability foundation across AWS, GCP, Azure, and on-premises, and the SLO/error-budget practices that keep the platform reliable. Forecast Engine: Guide the deterministic, multi-period capacity and cost simulation, its accuracy and reconciliation against actuals, and its evolution into an automated, always-on forecasting service. Future Capacity Reservation (FCR) automation: Guide the architecture that turns forecasts into reservation recommendations, how much capacity to reserve, in which providers and regions, and by when, aligned to hyperscaler procurement lead times. Thought Leadership & External Presence Represent ServiceNow at industry conferences and FinOps community events. Contribute to open-source projects and establish ServiceNow's presence in the modern data stack and FinOps ecosystem. Drive technical content creation including whitepapers, blog posts, and conference presentations. Build strategic relationships with technology vendors and the broader FinOps community. Collaboration & Integration Work autonomously with guidance from Engineering and FinOps leadership, owning the platform's technical direction. Partner deeply with the Senior Staff engineers who own each workstream, aligning their designs to one architecture without taking the keyboard away from them. Collaborate with DevOps, security, and platform teams on infrastructure, CI/CD, and compliance. Partner with product managers, FinOps practitioners, finance, and capacity-planning stakeholders to ensure the platform serves how the business actually plans, budgets, and governs cloud spend. Qualifications To be successful in this role, you have: Experience leveraging or critically thinking about how to integrate AI into work processes, decision-making, or problem-solving. 15+ years of experience in software or data engineering, with a track record of architecting and delivering large-scale, cloud-native, data-intensive platforms, with a Bachelor's degree; or 12 years and a Master's degree; or a PhD with 8 years of experience in Computer Science, Engineering, or a related technical field; or equivalent experience. Proven experience leading a large data platform migration or modernization off a legacy Hadoop or Cloudera stack (Impala, Hive, HDFS, Spark) onto a modern lakehouse, including inventory, schema and SQL translation, reconciliation against the source, cutover, and decommission of the old platform, ideally on a tight timeline. Experience architecting for regulated or government cloud environments (such as FedRAMP baselines or DoD Impact Levels) and designing systems that deploy across separate commercial and regulated boundaries with distinct isolation, data-residency, and compliance requirements. Deep expertise across the modern data stack (Trino/Presto, dbt, Apache Iceberg, orchestration) and in distributed-systems and cloud-native architecture. Hands-on experience designing streaming ingestion and change data capture, ideally log-based CDC from transactional databases such as PostgreSQL into a lakehouse, including schema evolution, delivery semantics, and reconciliation of streams with backfills. Working knowledge of streaming platforms (Apache Kafka or comparable) and the trade-offs in operating them, with the judgment to assess an incumbent and decide whether to keep or replace it under a deadline. Proven track record as the lead architect or top technical authority for a platform or program, setting direction that others build against. Strong systems and backend engineering depth, with the ability to go deep in any layer of the stack to make or unblock a hard technical decision. Demonstrated ability to operate at high velocity in greenfield environments with evolving requirements, shipping production-quality systems fast without sacrificing architectural integrity. Strong knowledge of data structures, algorithms, data modeling, design patterns, and performance optimization. Deep understanding of software quality principles including reliability, determinism, observability, security, and production r