Leveraging AI in automating Data Stewardship process
A large biopharma organization is using Informatica to create Prescriber master data based on IQVIA onekey / HCOS data, matching with HCP/HCO level transaction records from various internal / vendor sources. As any organizations handling MDM systems, major success criteria is to ensure the accuracy, semantic consistency and accountability of the Master data. Informatica uses certain fuzzy match capabilities to ensure these; however, requires significant human Data Stewardship to audit merge and publish requests. Here are the challenges faced.
- The Data stewards leverage several resources typically unavailable to HCM/MDM systems such as Google searches, external websites, and looping through corrected Names, addresses etc which makes the data stewardship process difficult to reproduce within the conventional MDM assets.
- Human intervention introduces a lot of subjectivity to the analysis and decision-making process due to varied skill levels and difficult to scale up as this can be transitioned only through ongoing activities.
- Stewards would spend considerable time on the simpler lookup patterns and resolving it, thereby resulting in inefficient utilization of the niche skills.
D Cube has taken all the challenges into consideration and customized our MDM AI solution which not only addresses above-mentioned challenges but also leverages advanced information retrieval, data enrichment and AI/ML capabilities to accurately replicate or optimize the manual stewardship process.
The solution simulates the stewardship process by accessing data from external websites that is unavailable to MDM and uses the enriched data to provide recommendations through AI/ML modelling. This also provides an easy and intuitive UI with unified view of the potential match records with all the enrichments and merge recommendations. This helps stewards confirm correctness of data before the decision making and select appropriate actions. The platform is also designed to integrate with the existing MDM solution.
The Outcomes / Impacts that MDM.ai created are listed below.
- Standardized lookups from external websites that is configured based on the attribute and lookup results availability helped implementing the standard operating procedure during lookup and analysis.
- AI/ML enabled automated system which mimics the data stewards mind to generate merge/no merge recommendations with a confidence score. ML algorithms take the subjectivity away.
- Empowerment by the AI/ML models, stewards with Niche skills can be re-assigned to other complex audit activities.
- AI/ML Model is frequently retrained based on the steward feedback would help accurately score recurring patterns. Stewards can focus on more varied scenarios during manual audits.
- Manual Data Stewardship activities were reduced by ~50% and time to onboard reduced from weeks to days.
Request a demo to find out how D Cube can help your organization gain productivity in the mastering process and automate the Data Stewardship activities.