Case Study Banner

CASE STUDY

An ML-powered automated emailing alert system assisted clinical trial disruption management for a global pharma leader

industry-iconCLIENT :Confidential
industry-iconINDUSTRY :Pharmaceutical
industry-iconDURATION :4 months
CLIENT :
Confidential
INDUSTRY :
Pharmaceutical
DURATION :
4 months

Business case

The global security monitoring team at a pharmaceutical major monitored geopolitical, socio-economic and natural calamities to ensure continuity of varied operations, manufacturing activities, and active clinical trials across geographies. With a complex drug development pipeline and long project plans, it notified dispersed teams of any impending risks that could cause operational delays.  

The global security team sends out detailed summary emails of all global events occurring across the world to all the teams within the organization. After which, the business global clinical supply chain team would manually go through these emails, figure out the priority and severity, and send action items emails to the relevant stakeholders within their purview.  However, manually processing these emails took a lot of time and was error prone. They also had historical data on past events and impact emails.

The global supply chain team wanted to eliminate the manual interventions and increase the efficiency of the email alerts process.  

Our Solution

After a thorough needs analysis, the i2e team developed an ML-powered algorithm that could better manage disruption alerts. Instead of manually sorting through the comprehensive emails, this model could read the comprehensive emails and parse data to identify alerts that could affect global clinical supply operations at a particular location.

By training the model on historical data, it predicted/forecasted the severity of the issue based on past historical trends.

Once the information is determined it is built into a pre-defined email template and sent to the relevant stakeholders. Our data analysts utilized the consolidated data to build real-time dashboards in Spotfire. 

Challenges overcome

  • Extracting data from unstructured data (emails), classifying and mapping them to bring a structure to the targeted emails.
  • Data Mining to parse the alert emails for obtaining clear picture of event severity and impact on operations. 

Benefits

  • The global clinical supply chain team could increase the efficiency of adverse events alert emails.
  • Targeted emails were sent to functional teams and right stakeholders facing the threat of disruption.
  • The dashboard gave the decision makers a comprehensive view of all the alerts sent along with their impact, severity and stakeholders

Results

Results