Business Impacts
Increased productivity for 45,000+ sales reps
Enhanced real-time responses
Improved data security and privacy
Customer Key Facts
- Country : Canada
- Industry : Healthcare Life Sciences
Problem Context
The client provides an AI-driven software as a service (SaaS) platform tailored for learning management in the life sciences industry, leveraging OpenAI ChatGPT 4 APIs. However, they encountered challenges such as outdated data, privacy concerns, and content reliability issues. To address these challenges, they turned to Quantiphi’s Generative AI SaaS, baioniq, hosted on AWS. The transition aimed to achieve a more efficient, updated, and secure solution.
Challenges
- Performance Monitoring: Ongoing monitoring of baioniq’s performance is essential to identify and address any issues that may arise, ensuring the system operates smoothly and meets user expectations.
- Onboarding new customers onto the baioniq platform and setting up APIs for mobile apps and Salesforce may require extensive resources and coordination.
- User Acceptance Testing (UAT) will be crucial to ensure that the new system meets the needs of diverse users, including representatives, managers, and specialists across various departments.
Technologies Used
Amazon CloudFront
Elastic Load Balancing
Amazon EKS
Amazon S3
Amazon Bedrock
Amazon Cloudwatch
Frontend Angular
Backend Node Server
ElasticSearch Storage,search & Logs
Kibana Data
Nat Gateway
Anthropic Claude v3
Solution
The solution entails replacing the Chat GPT 4-based assistant with Quantiphi's baioniq hosted on AWS, utilizing Anthropic Claude v3 - Sonnet, a powerful language model available in Amazon Bedrock. It addresses inquiries from sales representatives regarding medication dosage and side effects by analyzing uploaded documents. Users span across various roles including Representatives, Managers, MSLs/Clinical Specialists, across departments such as Sales, Marketing, CL&D, and Med Affairs.
During the Proof of Concept phase, baioniq is deployed, integrated with S3 and ElasticSearch, and connected to the client's environment via APIs. Management of licensing, performance monitoring, and User Acceptance Testing (UAT) will be overseen. Subsequent phases involve setting up APIs for baioniq in the client’s mobile app and Salesforce, onboarding customers, and conducting diverse UAT sessions with document uploads.
Results
- Enhanced Real-time Responses
- Increased Data Security and Privacy
- Optimized Platform Performance
