The Clinical Manufacturing Operating System
Eliminating the paper-to-production gap in orthotics
Peacocks Medical Group is building PEPS Ultra to fundamentally transform how custom orthotic devices move from clinical assessment to manufacturing. Today, a clinician’s expertise is trapped in handwritten notes, verbal instructions, and manual interpretation — a process that introduces delay, error, and dependency on highly specialised CAD technicians.
PEPS Ultra eliminates this paper-to-production gap entirely. By combining iPad-based 3D scanning, AI-powered voice interaction, and parametric CAD generation, it creates a seamless digital thread from the clinic chair to the factory floor.
“This is not just a software project. It is a fundamental change to how clinical expertise reaches the factory floor.”
Paper & Manual Process
Handwritten notes, verbal instructions, faxed prescriptions
iPad 3D Scan
Structure Sensor captures patient anatomy in seconds
Voice Command
Clinician speaks design intent, AI translates to parameters
Manufacturing Output
STL/STEP files ready for CNC, 3D printing, thermoforming
e.g. OPIE, Cliniko
Fatal Gap
Zero 3D capability. Cannot translate clinical intent into manufacturing specifications.
e.g. Autodesk, Materialise
Fatal Gap
Zero clinical context. Requires highly trained CAD technicians to interpret prescriptions.
Fragmented
Orthopedic Data
Manual
Interpretation
Specialised
CAD Technicians
Slow &
Error-Prone
We bring direct experience across the three pillars this project demands.
In-house R&D experience with depth sensors and spatial capture. We understand the scanning pipeline, accuracy considerations, and hardware integration — most development partners would need to outsource this.
We build cloud-native, multi-tenant systems as our core business. Patient management, order tracking, role-based access, and clinic portals are well within our delivery experience.
We build agentic AI systems that use tool-calling, natural language understanding, and multi-provider architectures. Directly applicable to the voice-driven design engine at the heart of PEPS Ultra.
Structure Sensor 3, the industry-standard medical-grade 3D scanner. Proven in orthopaedic clinics worldwide at $895 per unit.
Structure SDK licence: £500/month
Structure SDK 3.5.0 with built-in ML for foot landmark detection. 8 anatomical landmarks and 5 key measurements out of the box.
Leading commercial AI APIs with multi-provider failover. No custom ML model training required — leverages frontier models directly.
Open-source parametric CAD kernel (OpenCascade.js) compiled to WebAssembly. Manufacturing-grade B-Rep geometry in the browser.
Managed cloud services on GCP Cloud Run. Serverless containers that auto-scale to demand and operate on a pay-per-use model.
Vue.js 3 + Capacitor for cross-platform delivery. A single codebase serves the iPad native app and browser-based users simultaneously.
iPad-based scanning with automatic landmark detection. Medical-grade precision at 0.07mm accuracy.
Voice-driven device design. The clinician speaks, AI translates intent into parametric geometry in real time.
Server-side geometry engine producing manufacturing-ready STL and STEP output for CNC, 3D printing, and thermoforming.
Digital SOAP notes, prescription history, and scan archive with AI-powered semantic search across the full patient record.
Automated works orders and sales orders with live tracking dashboards, status notifications, and reporting.
Appointment management, orthotist scheduling, and session tracking across multiple clinic locations.
Tenant-isolated data architecture supporting multiple clinics and organisations with independent configuration.
Full scanning capability without internet connectivity. Local cache with automatic background sync when connection restores.
3D patient scanning using Structure Sensor 3. Medical-grade scans with automatic anatomical landmark detection. Works fully offline.
Full clinical workflow: patient records, voice-driven design, 3D preview, and order sign-off.
Browser-based access for manufacturing, order management, admin, and reporting. No install required.
| Component | Technology | Why |
|---|---|---|
| Clinician App | Vue.js 3 + Capacitor | One codebase for iPad app + browser |
| Scanning App | Native Swift | Direct hardware access to Structure Sensor 3 |
| 3D Viewer | Three.js | Renders foot scans and orthotic previews |
| API Backend | Node.js (Express) | REST API, auth, business logic |
| Database | MongoDB Atlas | Flexible document model, hosted on GCP |
| AI Agent (Primary) | Anthropic Claude API | Intent extraction + CAD script generation |
| AI Failover | OpenAI GPT / Google Gemini | Auto-failover, no vendor lock-in |
| Speech-to-Text | OpenAI Whisper API | Converts voice commands to text |
| Tool Integration | MCP (Model Context Protocol) | AI agent calls CAD tools directly |
| CAD Engine | OpenCascade.js (WASM) | Full parametric CAD kernel in Node.js |
| Cloud | Google Cloud Platform | UK-hosted (London), Healthcare API |
| Security | AES-256 + TLS 1.3 + MFA + RBAC | End-to-end encryption |
Opens PEPS Ultra, reviews schedule, accesses patient record. AI Ambient Scribe begins.
2-3 photos captured. CV overlays skeletal wireframe. 3D mesh + 80+ measurements.
Selects device category. Rules Engine + Logic Matrix. AI presents recommended starting point.
Voice commands modify geometry in real-time. Pricing adjusts automatically.
Reviews product spec, clinical note, pricing, 3D design. Taps “Finalize.”
Works Orders to factory, Sales Orders to SAGE, tracking activated, patient SMS sent.
10 minutes from patient greeting to manufacturing-ready output. Zero paperwork. Zero CAD skills required.
Across managing orders, production, and supply chain
Through rules engine and automation
Quality target for all devices
Down from 21+ days. 24 hours for insoles
Patient greeting to manufacturing-ready output
Clinicians describe what they need in natural language
| Requirement | Specification | Rationale |
|---|---|---|
| Data Security | AES-256 (rest) + TLS 1.3 (transit), MFA, RBAC | End-to-end encryption for all patient data |
| Interoperability | FHIR / HL7 Standards, HSCN | NHS hospital database integration pathway |
| Cloud Hosting | GCP London Region, UK data residency | UK data sovereignty compliance |
| Compliance Targets | ISO 27001, Cyber Essentials Plus, ISO 13485 | UK healthcare and medical device standards |
| Audit Trail | Full logging of all AI modifications + clinical actions | Explainable AI, regulatory traceability |
| Resilience | Multi-provider AI failover, multi-region GCP, 99.9%+ uptime | Always-on clinical access |
Foot scan → mesh + measurements
Full loop: scan → voice → AI → 3D → approve
Clinic-to-factory loop closed
TRL 5
Total ~40 weeks / 9 months · Phase overlaps maintain velocity · Target: TRL 5 by Q4 2026
Milestone
Clinician scans a patient’s foot on iPad and views the captured 3D mesh with extracted measurements.
Milestone
Clinician scans a patient, gives voice commands to design a bespoke orthotic, previews in 3D, and approves — all from an iPad.
Milestone
Approved design automatically generates a works order with manufacturing files, BOM, and delivery tracking.
Milestone
TRL 5 — Platform validated in a relevant clinical environment.
| Phase | Duration | Monthly Rate | Months | Phase Total |
|---|---|---|---|---|
| Phase 0: Discovery & Scoping | 2-4 wks | — | — | £15,000 |
| Phase 1: Foundation & Scanning | 10 wks | £20,000 | ~2.5 | £50,000 |
| Phase 2: AI Engine & Clinical Workflow | 15 wks | £20,000 | ~3.5 | £70,000 |
| Phase 3: Manufacturing & Orders | 13 wks | £20,000 | ~2 | £40,000 |
| Phase 4: Validation & TRL 5 | 11 wks | £20,000 | ~1 | £20,000 |
| Total Project Investment | ~40 wks | — | ~9 | £195,000 |
Phase 0 invoiced on commencement (£15,000 fixed scope)
Phases 1-4 invoiced monthly in advance at £20,000/month
Payment terms: Net 30 days from invoice date
Monthly steering reviews with milestone tracking and financial reporting
Investment figures are indicative and subject to confirmation following completion of Phase 0: Discovery & Scoping.
Final pricing will be agreed in the Statement of Work before Phase 1 commences.
Structure SDK licence: £500/month (required for 3D scanning capability)
Comprehensive delivery with transparent scope and no hidden costs.
Full design, development, and delivery of PEPS Ultra across all four sub-systems
Project management, sprint planning, and bi-weekly demo cycles
UX/UI design for all clinical and administrative interfaces
Quality assurance and automated testing throughout
Documentation for EIC Accelerator TRL 5 evidence package
Post-launch support during Phase 4 clinical pilot
AI tooling and infrastructure (Claude API, Whisper, GCP)
Weekly progress reporting and monthly steering committee updates
2-week sprint cycles with demo at end of each sprint
Weekly project team reviews
Monthly steering committee — milestones, financials, risk
Phase gate decisions before each new phase
Strict sprint planning, change control board, clear MVP definition.
Early clinician involvement in design, phased rollout with champions.
Early API discovery in Phase 0, dedicated integration sprints.
Real clinical scenario testing, clinician validation gates, override capabilities.
Proposal review and Q&A session with Vindico and Peacocks leadership teams
Commercial terms agreement and Statement of Work sign-off
Phase 0 kickoff — Discovery & Scoping begins
Phase 0 completion and Phase 1 gate decision
Ready to build what matters? Let’s talk.
Get in TouchUse landscape orientation with background graphics enabled