Required stack
필요 기술
PythonTensorFlowPyTorchscikit-learnFastAPIFlaskDockerKubernetesReactAngularFlutterReact NativeSwiftKotlinElectron.NET MAUISQLNoSQLAWSGCPAzureGitREST APIGraphQLCI/CD
Project brief
프로젝트 내용
I’m building an end-to-end healthcare application that blends AI-driven diagnostics and analysis, robust patient-management features, and forward-looking predictive analytics. The system must learn from three primary data streams—electronic health records, medical imaging, and genetic information—while remaining HIPAA-compliant throughout the pipeline.
The finished solution has to feel seamless on every major channel: a responsive web interface, companion mobile apps (iOS and Android), and a lightweight desktop client for Windows and macOS. Whichever framework you prefer—React or Angular for the web, Flutter, React Native or Swift/Kotlin for mobile, Electron or .NET MAUI for desktop—choose what lets you move fastest without sacrificing reliability.
On the back end, I expect modern ML tooling (Python, TensorFlow or PyTorch, scikit-learn, FastAPI/Flask for service layers) and clean DevOps practices (Docker, Kubernetes, CI/CD) so that the models stay reproducible and easy to update. Data pipelines should support structured EHRs, DICOM images, and VCF/BAM genomic files, and expose well-documented REST/GraphQL endpoints for future integrations.
Deliverables
• Data ingestion & preprocessing pipeline covering EHR, imaging, and genomics
• Trained and validated ML/Deep-Learning models for diagnosis, patient-risk scoring, and outcome prediction
• Unified web, mobile, and desktop front ends consuming a common API
• End-to-end security layer (encryption, audit trails, role-based access) meeting HIPAA guidelines
• Automated testing suite plus deployment scripts and clear developer documentation
Acceptance criteria: models reach agreed-upon performance metrics on a held-out test set, UI is fully responsive across devices, core workflows execute in under three seconds, and code passes all unit/integration tests.
If you can take the project from architecture to production launch, outline your approach, relevant past work, and timeline.
The finished solution has to feel seamless on every major channel: a responsive web interface, companion mobile apps (iOS and Android), and a lightweight desktop client for Windows and macOS. Whichever framework you prefer—React or Angular for the web, Flutter, React Native or Swift/Kotlin for mobile, Electron or .NET MAUI for desktop—choose what lets you move fastest without sacrificing reliability.
On the back end, I expect modern ML tooling (Python, TensorFlow or PyTorch, scikit-learn, FastAPI/Flask for service layers) and clean DevOps practices (Docker, Kubernetes, CI/CD) so that the models stay reproducible and easy to update. Data pipelines should support structured EHRs, DICOM images, and VCF/BAM genomic files, and expose well-documented REST/GraphQL endpoints for future integrations.
Deliverables
• Data ingestion & preprocessing pipeline covering EHR, imaging, and genomics
• Trained and validated ML/Deep-Learning models for diagnosis, patient-risk scoring, and outcome prediction
• Unified web, mobile, and desktop front ends consuming a common API
• End-to-end security layer (encryption, audit trails, role-based access) meeting HIPAA guidelines
• Automated testing suite plus deployment scripts and clear developer documentation
Acceptance criteria: models reach agreed-upon performance metrics on a held-out test set, UI is fully responsive across devices, core workflows execute in under three seconds, and code passes all unit/integration tests.
If you can take the project from architecture to production launch, outline your approach, relevant past work, and timeline.