Case Study
Healthcare Technology Company Optimizes Clinical Decision-Making with Generative AI
Industries
About the Customer
Our client offers a cloud-native care coordination platform that modernizes discharge and transition planning by digitizing paper-heavy workflows. By connecting patients, case managers, and medical vendors on a unified interface, the software optimizes provider utilization and streamlines administrative tasks. This integration drives a measurable reduction in both hospital readmission rates and average length of stay. On the backend, robust analytics deliver actionable insights and trend tracking needed to eliminate operational waste and mitigate clinical risks.
Challenge
Our Customer’s Key Business Objectives
The Healthcare Technology Company required a secure and scalable AWS infrastructure to support an automated, AI-driven matching engine for post-discharge care providers.
Their key goals included:
Key Amazon Web Services Used
- AWS Lambda
- Amazon API Gateway
- Amazon Simple Storage Service (S3) Vectors
- Amazon Bedrock
- Anthropic Claude 3.5 Sonnet (or Claude 4.5 Sonnet)
- Amazon Titan Text Embeddings v2
- Amazon DynamoDB
- Amazon Elastic Compute Cloud (EC2)
Third-Party Tools
- MongoDB
- GitLab
- Streamlit
Solution
How Cloudelligent Accelerated Healthcare Technology Company’s Objectives
A secure, AI-powered provider recommendation system was designed and implemented, transitioning our customer from manual, unstructured record reviews to a highly efficient, event-driven retrieval-augmented generation (RAG) architecture on AWS.
1
AI-Driven Semantic Search and Indexing
A robust knowledge base was established by consolidating fragmented datasets into a structured vector store. Amazon Titan Text Embeddings v2 was utilized to transform provider metadata, such as services and insurance, into high-dimensional vectors stored in Amazon S3. This enabled hybrid search capabilities, allowing the system to perform precise semantic matching and metadata filtering to ensure patients are paired with the most clinically appropriate providers.
2
Automated Orchestration and Retrieval Pipeline
A streamlined, event-driven pipeline was built using AWS Lambda to orchestrate the end-to-end recommendation logic. When a patient payload is received via Amazon API Gateway, the system triggers a multi-stage workflow to extract clinical intent, perform semantic retrieval, and utilize Anthropic Claude 3.5 Sonnet for intelligent re-ranking. This architecture replaces time-intensive manual analysis with a repeatable, scalable process that delivers results in seconds.
3
Clinical Reasoning and Intent Extraction
To resolve inconsistencies in patient data interpretation, advanced Generative AI logic was implemented via Amazon Bedrock. By leveraging Claude’s reasoning capabilities, the system analyzes unstructured clinical notes and social determinants to derive actionable insights. This ensures that every recommendation is backed by structured logic and clinical relevance, providing a standardized approach to decision-making that was previously prone to human variability.
4
Testing Enablement and Operational Handover
To support rapid iteration and product readiness, a lightweight UI was deployed on Amazon EC2 using Streamlit, allowing non-technical stakeholders to conduct User Acceptance Testing (UAT) and validate clinical outputs. The delivery cycle was further optimized by managing the codebase through GitLab and by implementing Amazon DynamoDB for comprehensive session logging and audit tracking. This ensured a seamless transfer of knowledge and a production-ready environment for the customer’s team.
Results & Benefits
Scalable, Accurate, and Clinically Validated Provider Matching
Cloudelligent’s solutions provide the customer with a production-ready MVP that bridges the gap between fragmented patient data and actionable clinical insights through a secure, automated Generative AI framework. Key outcomes include:
Enhanced Operational Efficiency and Automated Mapping
An automated patient-to-provider mapping system was implemented, eliminating manual lookups entirely. The transition to a serverless architecture enables care coordinators to receive 7 to 10 ranked provider recommendations instantly, significantly reducing manual effort and accelerating the discharge planning cycle.
Improved Clinical Accuracy through Generative AI Reasoning
The Healthcare Technology Company achieved a 4.4/5 overall evaluation score across clinical criteria while maintaining . By leveraging Generative AI reasoning and ranking, the system ensures every recommendation is aligned with patient needs, removing human variability and preventing suboptimal care coordination.
Scalable Architecture for Market Readiness and Growth
A production-ready MVP was delivered with the capacity to scale as additional provider datasets are integrated. This established a reusable framework for data ingestion and Generative AI insights, positioning our customer for a successful market launch and future Phase 2 enhancements such as AWS HealthLake and real-time data feeds.
Why Amazon Web Services?
Why Cloudelligent?
Ready to Embark on an Epic Cloud-Native Journey?
ascend as an industry pioneer with Cloudelligent
right by your side.
