Senior Operations BI Analyst

Writing complex multi-table joins, aggregations, window functions, stored procedures, and CTEs to extract, transform, and clean data from our ERP and data warehouse for specific operational analyses. You'll also review junior analysts' SQL.

Designing, building, and maintaining complex, interactive dashboards and reports for various operational teams. This includes advanced calculations (DAX/LODs), data modelling within the tool, and optimising performance. You'll also mentor others on best practices.

Performing advanced data blending, transformation, statistical analysis, and building predictive models for operational forecasting or anomaly detection. You'll use it for tasks too complex for SQL or standard BI tools.

Connecting BI tools to existing tables, writing custom queries directly against the data warehouse, and understanding the underlying data schema. You'll also be involved in optimising data models for performance and usability.

Writing custom queries directly against ERP tables to extract raw operational data. You'll need a deep understanding of the underlying data schema, key transaction codes (T-codes), and interdependencies between modules (MM, PP, SD) to validate your analysis.

Connecting to source systems (like ERPs) to build basic to intermediate process models. You'll use these to identify bottlenecks, deviations from ideal processes, and quantify the impact of process improvements in areas like order-to-cash or procure-to-pay.

Managing your project backlogs, creating and updating documentation templates for your workstreams, and facilitating requirements gathering sessions with operational stakeholders. You'll use it to keep projects organised and transparent.

• Plant Managers & Production Supervisors (they'll use your insights daily)
• Supply Chain Planners (for demand forecasts and inventory optimisation)
• Warehouse Operations Managers (to improve picking, packing, and shipping)
• Quality Control Teams (to analyse defect rates and root causes)
• IT & Data Engineering (for data source issues and new data pipelines)
• Finance Business Partners (to quantify the financial impact of operational changes)

• Logistics Providers (occasionally, when analysing carrier performance)
• Technology Vendors (if we're evaluating new BI or process mining tools)

Process mining is moving beyond just 'seeing' what happened to 'predicting' what will happen and 'prescribing' what to do. As a Senior Analyst, you'll need to move from interpreting basic process maps to building more complex models, identifying root causes of deviations, and even simulating potential process changes to quantify their impact before implementation.

Conformance checking · Predictive process monitoring · Process simulation for 'what-if' analysis · Automated root cause analysis in process mining

• This week: Review advanced tutorials or webinars on Celonis or UiPath Process Mining, focusing on conformance checking.
• This month: Identify a specific operational process (e.g., order fulfilment) and build a more detailed process model, including variants and deviations.
• Month 2: Work with an operational team to identify a specific process bottleneck and use process mining to quantify its impact and suggest solutions.
• Month 3: Explore the simulation capabilities within your process mining tool (or a separate simulation tool) to model a proposed process change.

Quick win: Use process mining to visualise the 'happy path' vs. 'exception paths' for a core operational process and present the findings to the relevant team. They'll be surprised by what they see.

While you might not be a full-blown Machine Learning Engineer, as we deploy more predictive models (e.g., for demand forecasting, anomaly detection, predictive maintenance) into production, you'll need to understand how to monitor, maintain, and retrain them. It's not enough to build a model; you need to ensure it keeps performing well in the real world.

Model monitoring and drift detection · Automated retraining pipelines · Version control for models and data · Explainable AI (XAI) for operational transparency

• This week: Read an introductory article or watch a video series on MLOps concepts.
• This month: Identify one existing predictive model we use (e.g., a demand forecast) and understand its current monitoring and retraining process.
• Month 2: Propose a small improvement to how we monitor that model's performance, perhaps a new metric or alert.
• Month 3: Work with Data Engineering to implement a basic version of your proposed monitoring improvement.

Quick win: Start tracking the actual vs. predicted values for one of our operational forecasts in a simple dashboard. This is the first step to understanding model drift.