Pharma due diligence automation

Implementation & training

Our team set up Enate to manage the medicine registration flow and orchestrate tasks between employees and RPA bots. We also configured UiPath robots and leveraged GenAI capabilities to validate dossier structure and content against predefined rules as well as collect medicine data from product trackers and automatically distribute it to relevant stakeholders via email.

Before the full-scale deployment, we released a pilot version that automated the entire due diligence process for one region with a limited number of compliance checks and business notification process management. This approach helped us ensure stakeholder alignment, reduce the risk of rework, and establish effective communication processes. During the pilot stage, the team finalized functional requirements and formalized dossier due diligence business rules as well as configured bot- and ML-based compliance checks and set up integrations with other corporate systems, laying the foundation for the platform’s scalability.

During full-scale implementation, we configured the necessary rules to assess the compliance of all modules within a dossier, incorporating around 500 regulatory rules. We also automated due diligence and notifications across other company branches, configuring Enate workflows to meet the regulatory requirements of 50+ countries.

After the deployment, we conducted a series of stakeholder workshops to demonstrate the system's functionality and onboard users to the new automated workflows, while addressing questions as they arose.

Functionality

Dossier due diligence

Comprehensive checks of the required document availability and content within each dossier are performed by the UiPath bot in combination with Generative AI. UiPath bots, built according to the customer’s business rules, can be run at any stage of the submission process. Thanks to the system configurations, country- or dossier-specific rules can be easily modified or added, which facilitates the expenditure of the solution’s automation capabilities.

The actual tasks for the bot to run automated checks are enabled with the help of Enate, with three main processes involving registration (initial registration of a medicine), variation (post-registration changes), and renewal (re-registration after the initial registration expires).

RPA enhancement through Gen AI & ML

To support due diligence involving complex decision-making and processing heterogeneous, unstructured data, we enhanced RPA-based compliance checks with Generative AI. We combined a UIPath-powered automation solution with Large Language Models (LLMs) to enable context-aware data analysis and intelligent automation.

We opted for the GPT model to work alongside UiPath bots, allowing them to process various document templates and extract, analyze, and structure document data. By interpreting document context, the GPT model eliminates the need for strict bot rules and enables prompt-driven dossier due diligence. The interaction between the GPT model and UiPath bots occurs via API calls.

First, the UiPath bot segments customer documents, including scans and non-editable PDFs, into logically consistent sections to ensure accurate contextual interpretation of each section during processing, even for very large files. Then, the bot sends specific document sections to the GPT model together with prompts and automatically captures, structures, and records the responses. After that, consolidated compliance check results, including detailed findings and required follow-up actions, become available in Enate for review and management.

For complex content analysis, such as molecular formulas’ structure checks, we developed a custom Python-based ML model triggered by the UiPath bot.

Business notifications automation

A UiPath bot automates the preparation and distribution of regulatory notifications by gathering data from multiple systems, compiling it into the required format, attaching supporting documents, and sending notifications to stakeholders. This helped reduce processing time from around 30 minutes to 1-2 minutes, while still allowing human intervention if information is missing or requires adjustment.

Additional functionality

Task management

Enate supports end-to-end task management, including project and project template creation, task assignment between employees and RPA bots, and progress tracking.

We also customized the Enate solution’s interface, developing custom UI components to support verification process validation, case-specific data visualization, and dynamic field control.

User enablement & security

Enate enables role-based access control, allowing multiple users to collaborate on the same dossier while ensuring secure user authentication and permission management.

Reporting & analytics

Managers, scientists, and other stakeholders can monitor the dossier due diligence process via Enate dashboards. To address the customer’s specific reporting needs, we extended Enate with custom fields to track additional metrics, including error rates, error occurrence times, automation process duration, and verification success rates. While Enate provides broader analytics and reporting capabilities, UiPath bots generate detailed due diligence check reports.