SENIOR LOOKER / BI ENGINEER

SENIOR LOOKER / BI ENGINEER GCP Data Platform | Client engagement | 12 months, ~720 hours | Remote (India) ABOUT THE ENGAGEMENT Client has moved its data warehouse from Oracle to BigQuery and is engaging Searce to (a) stand up Looker as the governed consumption layer for the business, (b) bring BigQuery costs and performance under control, and (c) deliver a conversational analytics agent over the governed data. One senior engineer owns this end to end. The client has been explicit that Looker is the primary requirement and GCP data engineering the secondary one; a candidate strong on BigQuery but light on LookML will not clear their interview. WHAT YOU WILL OWN Primary — Looker (roughly 60–70% of the role) • Set up and administer the Looker instance: BigQuery connections, users and groups, roles and permission sets, folder structure, content governance and housekeeping. • Design and build the LookML semantic layer on curated Gold-layer BigQuery tables: views, explores, dimensions, measures, joins, derived tables and PDTs, with aggregate awareness and partition-aware SQL to keep queries fast and cheap. • Design and deliver business dashboards and reports prioritised with client stakeholders; iterate on adoption, not just delivery. • Implement access grants and user attributes for row- and column-level security across business units and regions. • Tune Looker performance: query analysis via System Activity, PDT strategy, caching and datagroups, and pushing heavy logic down into BigQuery. • Drive user enablement and rollout: training, documentation, office hours, and measuring dashboard usage. • Build and deploy the Conversational Analytics Agent over governed Looker/BigQuery datasets: semantic-layer configuration, prompt design, guardrails, response validation against curated data, and UAT. Secondary — GCP data engineering (roughly 30–40%) • Review and optimize BigQuery: partitioning, clustering and incremental refresh on the top pipelines; audit expensive jobs via INFORMATION_SCHEMA; set user query limits and cost guardrails. • Operate Data Stream CDC pipelines to a <15-minute freshness SLA; build and run Airflow / Cloud Composer DAGs with monitoring, alerting and failure recovery. • Shape the Bronze/Silver/Gold data model so ingestion is decoupled from reporting; implement basic data quality checks and alerts. • Administer the toolchain (BigQuery, Data Stream, Composer, Looker) and produce monthly cost and health reports. MUST-HAVE • 2+ years hands-on with Looker (Google Cloud's enterprise BI tool, not Looker Studio) in a production setting, with demonstrable LookML ownership: you have written explores, views, measures, derived tables and access grants yourself and can walk through them in an interview. • Experience administering a Looker instance: connections, users/roles, folders, content governance, System Activity. • Strong SQL and 1+ years on BigQuery, including partitioning, clustering, cost analysis and query tuning. • Working knowledge of a GCP orchestration stack: Airflow / Cloud Composer, plus at least one of Data Stream, Dataflow or Pub/Sub. • Comfortable working directly with business stakeholders and a client CIO/IT team; able to run requirement sessions and demos independently. NICE TO HAVE • Looker LookML Developer certification; Google Professional Data Engineer. • Experience with Gemini in Looker, Conversational Analytics, or Google Agentspace. • Oracle background and exposure to media, publishing or advertising data. • Python for pipeline and automation work; Git-based LookML development workflows. SCREENING NOTE FOR RECRUITERS AND INTERVIEW PANEL • Ask for a specific LookML artefact the candidate built (explore, PDT, access grant) and have them explain the design choice. Candidates who describe Looker Studio blending, BI Engine or Custom SQL sources are answering a different question. • Weight the technical round roughly 60% Looker / LookML, 40% BigQuery and pipelines, in line with the client's stated priority. • Candidates should be able to speak to Looker adoption outcomes (usage, stakeholder rollout), since dashboard adoption is a Full Timeual success criterion on this engagement