Senior Power BI Engineer

We are building a centralised marketing and sales analytics platform for a large multi-market consumer brand: a governed BigQuery data layer feeding a single, certified Power BI reporting product used by head-office marketing and sales teams and by every national market. It replaces a fragmented estate of hand-built, per-market dashboards and manual spreadsheet exports with one version of the truth.

You will own the Power BI half of that platform end to end — the semantic model, the KPI logic, every report page, and the executive PDF report generated from it. This is not a dashboard-assembly role. One model and one report template must serve every market through filtering, hold up under executive scrutiny, and reconcile to source platforms on demand.

Experience: 6+ Years

Location: Kolkata (Onsite)

‍Key Responsibilities:

‍

Strategic Marketing & Brand Building

• Design, build and own the certified Power BI semantic model over a governed star schema in the cloud warehouse — relationships, cardinality, role-playing dates, hierarchies and field-level documentation.
• Choose and justify the connectivity approach (Import, DirectQuery or composite), and implement incremental refresh, aggregations and partitioning to keep refresh windows and query response inside agreed service levels.
• Author the KPI layer in DAX: a frozen, documented catalogue of measures covering traffic, engagement, conversion and funnel performance, with consistent year-on-year, year-to-date, quarter-on-quarter and custom period-comparison logic.
• Implement weighted and normalised benchmark measures so small and large markets can be compared fairly against a group average.
• Keep the model performant and maintainable — calculation groups, variables, disciplined measure structure, and regular tuning with Performance Analyzer, DAX Studio and Tabular Editor.

‍

Report and experience design

• Build the report pages the business actually asked for: a real-time KPI tracker with progress-to-target visuals, market-versus-group benchmarking, product-level performance, an end-to-end lead-to-order funnel with banded performance thresholds, channel and traffic-source analysis, and social performance.
• Design for flexible self-service: independent metric selection per visual, multi-market comparison, switchable chart types, and free choice of reporting and comparison periods — without users needing to export anything.
• Apply a consistent, accessible visual language across every page: clear labelling, sensible visual hierarchy, and charts rather than dense tables.
• Configure workspaces, apps, audiences and deployment pipelines across development, test and production, and implement row-level security where required.

‍

Executive reporting layer

• Build and automate the recurring executive PDF report generated directly from the platform, using paginated reports and export automation, so a market or period can be selected and a complete, correctly formatted pack produced in one action.
• Ensure exported visuals remain visuals — presentation-ready output, not screenshots of tables.

‍

Quality, reconciliation and trust

• Build data quality and anomaly surfacing into the reporting layer: freshness indicators, missing-period detection, and outlier alerting on unexpected drops or spikes.
• Reconcile platform figures against the underlying source platforms, document the variances that are structurally expected, and explain them clearly to non-technical stakeholders before they are discovered in a meeting.
• Write and maintain the model and report documentation, KPI definitions and handover material.

‍

Delivery

• Work to a gated, fixed-scope delivery plan with hard go-live dates, translating business requirements into agreed KPI definitions and report specifications.
• Run user acceptance testing on a reference market, manage a controlled review-and-revision cycle per report page, and support training and handover.

‍

‍Required Skills & Qualifications:

• 6+ years in business intelligence or analytics engineering, with at least 4 years hands-on in Power BI on production, business-critical reporting.
• Advanced DAX — time intelligence, calculation groups, context transition, performance-conscious measure design. You can explain why a measure is slow, not just that it is.
• Proven semantic model design over a dimensional warehouse: star schema, incremental refresh, composite models, aggregations, gateways and capacity considerations.
• Strong SQL, and practical experience querying a cloud data warehouse (BigQuery preferred; Snowflake, Redshift or Synapse considered) with an understanding of partitioning, clustering and query cost.
• Direct experience modelling digital analytics data — particularly GA4 event-level exports in a warehouse, including nested and repeated structures, sessionisation and conversion definitions. This is the single most important differentiator for this role.
• Paginated reports (Power BI Report Builder / RDL) and export automation for pixel-accurate PDF output.
• Report and dashboard design judgement — you can take a vague executive requirement and produce something that answers it in seconds, and you can defend your layout decisions.
• Power Query / M for shaping and troubleshooting, and comfort with source-control and deployment discipline.
• Confident written and spoken English, and the composure to present work directly to senior stakeholders and take feedback in the room.

‍

Desirable:

• Marketing, media or e-commerce measurement background — an understanding of what marketers and sales leadership actually need to see.
• Working knowledge of consent management and its effect on reported figures, and how modelled data behaves differently across interfaces and warehouse exports.
• Microsoft Fabric, Power BI Premium / Fabric capacity administration.
• Multi-market or multi-tenant reporting at scale, including row-level security design.
• dbt or an equivalent transformation framework; Python for supporting automation.
• Custom or advanced visuals (for example Deneb / Vega-Lite) where the standard library falls short.
• Relevant Microsoft certification (PL-300 or equivalent) — useful, but the portfolio matters more.