Chisom – BigQuery, SQL, Snowflake

The finance reporting layer for a podcast and audiobook streaming platform, built on BigQuery, dbt and Airflow. It produces the core revenue-transaction fact tables that every finance report depends on, and automates sales-invoice posting straight into Microsoft Dynamics 365 Business Central over REST API — removing manual invoice entry entirely, with idempotency and master-data sync so the two systems never drift.

• Built the finance reporting layer on BigQuery, dbt, and Airflow, including the core revenue-transaction fact tables used across finance reporting.
• Designed and built a BigQuery to Airflow to Microsoft Dynamics 365 Business Central REST-API pipeline to automate sales-invoice posting, with idempotency and master-data sync between systems.
• Served as the sole engineer on the engagement, handling requirements, modelling, orchestration, and stakeholder communication with the client finance team end to end.

The data platform behind a connected-audio consumer hardware company. The core piece reconciles retailer sell-out data against internal sell-in figures across wholesalers on multiple continents — heterogeneous export formats, no shared entity IDs, and numbers that disagree by default — and feeds sell-out forecasting plus media-mix models that attribute offline retail uplift to specific campaigns. Also delivered the reverse-ETL user-stats product that contributed roughly €350k of Q4 2025 revenue.

• Built the reverse-ETL user-stats product that contributed approximately €350k of Q4 2025 revenue.
• Led the wholesale sell-out pipeline reconciling retailer sell-out data against internal sell-in figures across wholesalers on multiple continents.
• Drove an architecture upgrade separating dev and prod databases, eliminating redundant dbt projects, and adding CI tests, incremental loading, and clustering.
• Led implementation of performance-marketing metrics and media-mix modelling to attribute offline retail sales uplift to specific campaigns.
• Defined event-naming templates and data contracts so new Toniebox 2 data flowed into existing models with no remodelling.
• Mentored 3 analytics engineers, set standards for data quality, and guided the team through ambiguous undocumented integration work.
• Optimised the agentic AI workflow that generated dbt staging models, directing and reviewing output for the team.

A scalable creator-revenue and payout data platform for podcast creators and audiobook publishers. Tested, versioned dbt models covering contractual revenue splits, minimum guarantees, and edge cases; a reconciliation layer aligning streaming, billing, and contract data before payouts. A monthly payout cycle reduced from 18 to 7 days, approximately €1.2M in annual savings, and a zero-downtime migration from SAS to BigQuery with around 70% lower warehouse costs.

• Automated the revenue-share and payout process for podcast creators and audiobook publishers, encoding complex contract logic as tested, versioned dbt models.
• Spearheaded migration from legacy SAS to BigQuery with no downtime and built the core models serving data scientists, analysts, and product analysts.
• Automated revenue reporting with outlier tests that flagged anomalies for finance deep-dives before payouts went out.
• Automated Customer Lifetime Value computation with a 5-year forecast and surfaced it via a self-serve Looker dashboard.

The analytics platform for one of Europe's largest retail brokerages. Airflow-orchestrated ingestion from external APIs into Snowflake and S3, feeding the models behind regulated financial reporting. Automated recurring reporting to cut manual data processing ~50% across finance, product, security, operations and strategy, and built the company-wide KPI dashboard leadership used to set targets and report to the board.

• Used Airflow to orchestrate ingestion from multiple external APIs into the data lake and warehouse using Snowflake and AWS S3.
• Built unit tests, scheduling, and shared Python tooling.
• Automated reporting to reduce manual data processing by approximately 50% for finance, product, security, operations, and strategy teams.
• Helped build an internal A/B-testing platform and the company-wide KPI dashboard used by leadership to set targets and report to the board.
• Mentored more than 100 business users on Looker, Snowflake, and SQL, promoting best practices in analysis and visualisation.

Analytics for a banking-as-a-service platform. Automated B2B invoicing to cut issue time from 7 business days to 2, speeding remittance, and refined default-probability models used in credit risk.

• Automated invoicing to cut issue time from 7 to 2 business days, speeding B2B remittance.
• Ran SQL and analytics training for business users.
• Refined default-probability prediction models, improving accuracy.