MVP Estimates & Implementation Roadmap¶

Document Purpose: This document provides the engineering effort estimates (in man-hours) and the sequenced implementation roadmap for the MVP Phase 1.

1. MVP Effort Estimates (Man-Hours)¶

Summary: - Total Estimated Effort: 750 Hours - Team Size Assumption: 3 Developers (1 Full Stack, 1 Backend/Data, 1 AI/ML/Data) - Timeline Estimate: ~9 weeks (assuming full-time)

Phase	Module	Components	Frontend (Hrs)	Backend (Hrs)	Data (Hrs)	QA (Hrs)	Total (Hrs)
Phase 1	0. Infrastructure	VPC, RDS, S3, LB (INF-001)	0	40	10	0	50
Phase 1	1. Foundation	Setup, Auth, RBAC (SEC-001)	16	24	8	8	56
Phase 1	2. Ingestion	HL7, PACS, File Watchers (ING-001)	0	16	64	16	96
Phase 1	3. Core Data	Patient API, Search, Vitals (API-001)	24	32	16	12	84
Phase 1	4. Timeline UI	Chronological View, Filters (UI-003)	40	8	0	8	56
Phase 1	5. Clinical UI	Labs, Imaging, Reports Views (UI-004)	32	16	8	8	64
Phase 1	6. Oncology	Chemo, TNM Staging, RECIST (ONC-001)	32	24	16	12	84
Phase 1	7. Notes & Scribe	Pipeline, Annotation UI (PROC-005)	16	24	16	8	64
Phase 1	8. AI/ML Engineering	Model Training & Inference (ML-001)	0	24	80	16	120
Phase 1	9. Interop	FHIR R4 Generation (API-004)	0	40	8	8	56
Phase 1	10. Alerts	Rule Engine, Notif System (NOTIF-001)	8	8	0	4	20
	Totals		168	256	226	100	750

2. UI Achievement & Progress¶

This section tracks the progress of the UI/UX development against the budgeted engineering hours. It distinguishes between Achieved Design (Mockups) and the remaining Technical Hardening.

Phase	Capability	Total UI Budget (Hrs)	Achieved (Design)	Status
Phase 1	Timeline & Core Navigation	40	32	[UI: Complete]
Phase 1	Clinical Overview & Vitals	32	24	[UI: Complete]
Phase 1	Oncology Staging & Treatment	32	24	[UI: Complete]
Phase 1	Notes & Scribe UI	16	12	[UI: Complete]
Phase 1	Search & Foundation UI	40	28	[UI: Complete]
Phase 1	Alerts & Interop UI	8	0	[UI: Pending]
Phase 2	Clinical Decision Hub (AI Regimens)	160	0	[UI: Deferred]
Phase 2	Research & Population Analytics	120	0	[UI: Deferred]
Phase 2	Surgical OT Worklist & Case Log	150	0	[UI: Deferred]
Phase 3	GNN Confidence & Outcome Prediction	200	0	[UI: Deferred]
Phase 3	Knowledge Explorer & Causal Flow	180	0	[UI: Deferred]
	Totals	1018	120 (~12%)

[!IMPORTANT] Hardening & Integration Buffer: Approximately 25-30% (50 Hours) of the total frontend budget remains to be invested. This effort is dedicated to API Integration, Responsive Layouts (iPad/Mobile optimization), and Verification Testing to ensure mockup fidelity in the live production environment.

3. Iteration List (Dependency Based)¶

This roadmap orders use cases by technical dependency: Data Layer → Backend Services → Frontend UI → Functional logic.

Iteration 0: Cloud Infrastructure (Week 1)¶

Goal: Provision the environment.

[INF-001] Infrastructure Provisioning: Setup VPC, Subnets, RDS, S3, and EKS/Compute clusters.

Iteration 1: Data Ingestion & Foundation (Weeks 2-3)¶

Goal: Get data IN and secure the platform.

[SEC-401a] Authentication & RBAC: Setup Auth0/Cognito, User roles (Doctor/Admin).
[SEC-002] Audit Logging: Middleware to log every data access.
[ING-001] HL7 ADT Listener: Accept patient demographics and admission feeds.
[ING-006] PACS JSON Feed: Parse imaging metadata streams.
[ING-004] Lab Results (ORU): Parse numeric lab values and reference ranges.
[API-001] Patient CRUD: Basic GET/POST patient profile APIs.

Iteration 2: Core Read Services & Basic UI (Weeks 4-5)¶

Goal: Display the 'Skeleton' of the patient record.

[API-005] Global Search: Elasticsearch indexing of patients.
[UI-001] Patient List: Searchable roster UI.
[UI-002] Patient Overview: Demographics, Vitals sparklines header.
[API-004] FHIR Transformation: Backend service to map DB schema to FHIR R4 (needed for structured data).
[UI-004] Labs Dashboard: Tables and trend charts for Hematology/Chemistry.

Iteration 3: Timeline & AI/ML Training (Weeks 5-6)¶

Goal: The "Killer Feature" - unifying all events with AI assistance.

[API-002] Unified Timeline API: Aggregating Labs, Imaging, Notes into one time-series.
[UI-003] Timeline Component: High-performance vertical list with event grouping.
[ML-001b] AI Model Fine-tuning: LoRA/QLoRA training for Whisper (Scribe) and OCR tuning.
[OPS-302] Inference Optimization: GPU acceleration and performance tuning for real-time Scribe. 15b.[PROC-001] OCR Pipeline: Tesseract/PaddleOCR execution for hospital PDFs.
[UI-006] Document Viewer: Viewer for reports linked from timeline.
[PROC-005] Notes/Scribe Pipeline: Async job to process raw PDFs/Audio into structured notes.

Iteration 4: Oncology Specifics (Week 6)¶

Goal: Specialized cancer workflows.

[ONC-001] Diagnosis & Staging: UI for TNM staging and histology.
[ONC-020] Treatment Lines: Visualization of Chemo/Radio cycles.
[UI-006] Annotation: Allow doctors to tag timelines/docs.
[ONC-011] RECIST Display: Show tumor measurements from PACS.

Iteration 5: Polish & Interop (Week 7-9)¶

Goal: Production readiness.

[NOTIF-001] Basic Alerts: Rule engine for critical labs/missed appts.
[API-006] Presigned Uploads: Optimized file upload flow.
[INT-001] EMR Deep Link: "Context launch" from hospital EMR.
[QAS-001] Performance Tuning: Ensure <2s API response.

4. Deferred Vision Feature Estimates (Phase 2/3)¶

The following features were identified in the high-fidelity mockups but are deferred to subsequent phases. These estimates are preliminary and subject to detailed scoping.

Domain	Feature	Estimated Effort (Hrs)	Priority
Clinical Decisions	Clinical Decision Hub (AI Regimens)	160	Phase 2
Clinical Decisions	GNN Confidence & Outcome Prediction	200	Phase 3
Research	Population Search & Side-by-Side Comparison	120	Phase 2
Research	AI-Powered Clinical Trial Matching	140	Phase 3
Operations	Multilingual Support (Telugu/Regional)	80	Phase 2
Operations	Doctor Availability & Leave Management	60	Phase 2
Analytics	Causal Flow & Relational Knowledge Explorer	180	Phase 3
Surgical	Specialized OT Worklist & Intra-op Case Log	150	Phase 2

Parent Document: MVP Scope | Visual Gaps: Visual Gaps & UI Maturity Roadmap