Use Cases for GenAI Training Capstone Project
1. Clinical Guidelines Q&A Assistant
Problem Statement:
Doctors and healthcare teams need quick access to treatment guidelines for diseases such as
diabetes, hypertension, asthma, COPD, and oncology. A RAG assistant can answer guideline-based
questions with citations.
Data Requirements:
● Public clinical guidelines
● Disease-specific treatment protocols
● Drug and dosage reference material
Expected Output:
● Evidence-backed answer
● Relevant guideline section citation
● Treatment pathway summary
● Warning when more clinical judgment is required
2. Healthcare Knowledge Graph + RAG Assistant
Problem Statement:
Healthcare data includes relationships among diseases, symptoms, drugs, genes, biomarkers, and
clinical trials. A combined knowledge graph and RAG solution can answer complex biomedical
questions more accurately.
Data Requirements:
● Disease-drug-symptom relationship data
● Research articles
● Drug labels
● Trial registry records
Expected Output:
● Relationship-based answer
● Graph path explanation
● Source-backed evidence
● Visual relationship summary
3. Healthcare Insurance Prior Authorization Assistant
Problem Statement
Healthcare insurance prior authorization is a process where hospitals, providers, or patients need
approval from the insurance company before a treatment, surgery, diagnostic test, hospitalization, or
medication is covered.
Data Requirements:
● Covered treatments and procedures
● Non-covered services
● Exclusions
● Waiting periods
Expected Output:
● The GenAI assistant should generate a structured prior authorization summary.
4. Rare Disease Knowledge Assistant
Problem Statement
Rare disease information is often scattered across multiple sources such as medical websites,
research papers, case reports, genetic databases, patient advocacy documents, and clinical trial
registries.
Patients, caregivers, researchers, and healthcare teams may find it difficult to quickly understand:
● Disease symptoms
● Genetic causes
● Diagnostic tests
● Available treatments
● Specialist referral needs
Data Requirements:
● Rare disease overview documents
● Disease symptoms
● Causes and inheritance patterns
● Genetic mutation details
Expected Output:
● The assistant should provide structured, reference-backed answers.