Our Customers’ Success
We take great pride in making each and every one of our client engagements a success to ensure a 100% satisfaction rate.
Global Financial Institution
Client challenge: The client’s rule-based fraud detection on the payment processing system was ineffective. The rules were manually development by fraud experts by analyzing different fraud transactions, leading to cognitive biases, recency bias, and simplicity.
Our Approach: Leveraged machine learning to develop a fraud detection model. The algorithm was trained on 1 year of historical data consisting of billions of transactions with more than 200 variables.
Outcome: Implemented model was able to pinpoint 65% of the fraud transactions and reduce fraud-related costs by 73%.
Large medical practice
Healthcare Operations Dashboards
Client challenge: Physicians/staff, operations and C-suite lacked transparency into patients’ quality metrics, overdue patient paperwork and physician utilization.
Our Approach: Created an automated dashboarding solution that regularly auto refreshed by leveraging AWS to store EMR data, Alteryx to process and analyze the data and Tableau to host dashboards. Designed dashboards using an agile methodology with iterative stakeholder feedback to create the insights needed.
Outcome: Improved overall patient care through the transparency of quality metrics (58% to 71% overall quality metrics satisfied) and decreased time to bill and fill prescriptions to better serve patients (52% decrease in overdue paperwork).
Global Manufacturing Company
Client challenge: The client had a suboptimal part maintenance scheduling where maintenance was executed every 2 weeks, regardless of each component’s propensity to breakdown.
Our Approach: Leveraged stochastic models on each machine component to fit probability distributions of breakdown. Optimal maintenance intervals were determined based on lost production costs from breakdowns, maintenance costs and predicted breakdown probability.
Outcome: Machine breakdowns were reduced by 40% providing a $2.2M savings in lost production.
Global Pharmaceutical R&D Firm
Client challenge: COO had limited visibility into state of R&D portfolio, functions, projects and studies. As well as a need to understand especially projects that were behind schedule and over budget.
Our Approach: COO had limited visibility into state of R&D portfolio, functions, projects and studies. As well as a need to understand especially projects that were behind schedule and over budget. Created a data lake to integrate data from Planisware and SAP to quickly build a dashboarding solution in Tableau. The Tableau-based solution enabled the COO to quickly go from an entire portfolio-level view and drill down into individual functions, projects and studies.
Outcome: COO and R&D Operations team was able to quickly identify and intervene in problematic projects and studies. Severely over budget projects were reduced by 20%.
Global Banking Company
Client challenge: The client was looking to increase customer retention through a deeper understanding of transactions and interactions that led to customer attrition.
Our Approach: Leveraged 300 plus variables across customer interaction channels and transactions to segment customers into 16 distinct segments. The number of variables was reduced through correlation indexes. A random forest model was leveraged to understand highly correlated events with customer attrition within each customer segment.
Outcome: Determined banking branch visits and fees were highly correlated with customer attrition. Recommended elimination of fees for certain customer segments and incentivizing branch visits for customers at risk for attrition.
Global Toy Manufacturer
Product, Sales and Portfolio Management
Client challenge: The client had over 10,000 SKUs and no visibility into SKU-level profitability. The large number of SKUs also contributed to high working capital requirements to hold products in inventory. The challenge was to determine the optimal product portfolio to maximize revenue and reduce inventory costs.
Our Approach: Worked alongside IT to model and source the data required to build a suite of dashboards used by Sales, Marketing, Product and Finance that answered all immediate queries – the most important dashboard being the portfolio analysis dashboard. In addition we set them up for continued success with self-service access to the data sources we developed for analysts in each team to answer future questions.
Outcome: The organization is enabled with self-service access to a multitude of data sources and the new dashboards enabled management to quickly determine SKU profitability, sales and costs.
Decreased fraudulent cost by 73% with machine learning.
Improved patients meeting quality metric measures by 22%
Reduced machine breakdowns by 40% through preventative maintenance.
Reduced project budgets by 20%
Substantially increasing customer retention.
Enabled management to quickly determine SKU profitability.