Home
ProductOperational ImpactIT & Procurement
Publications
About UsNewsCareersMedia Kit
Contact Us Login Book a Demo

IT & Procurement

Everything your team needs to evaluate, approve, and deploy Us2.ai. Built for hospital IT committees, procurement officers, and security reviewers.

Predictable Deployment Process

Us2.ai is designed to fit into existing hospital infrastructure with minimal IT effort.

1

Scoping & Assessment

Week 1–2

Review infrastructure, DICOM routing, and integration points. Identify cloud or on-premise deployment path.

2

Installation & Configuration

Week 2–3

Deploy Us2.ai server (on-premise) or provision cloud instance. Configure DICOM routing and system connectivity.

3

Validation & Testing

Week 3–4

Run test studies, validate output against your existing workflow, and confirm integration with PACS/CVIS/EMR.

4

Go Live

Week 4+

Enable production routing. Ongoing monitoring and support. Can be disabled at any time without data migration.

Deployment Options

Choose the deployment model that fits your organization's security and infrastructure requirements.

Cloud

Managed cloud deployment. No on-site hardware required. Studies routed via secure DICOM connection. Automatic updates and scaling.

  • Zero hardware procurement
  • Automatic software updates
  • Elastic scaling for volume spikes
  • Data encrypted in transit and at rest

On-Premise

Dedicated hardware deployed within your data center. Data never leaves your network. Full control over infrastructure and access.

  • Data stays within your facility
  • Integrates with internal DICOM routers
  • Compatible with air-gapped environments
  • IT retains full infrastructure control

Minimum Hardware Requirements

Minimum specifications for on-premise deployment. Cloud deployment requires no hardware procurement.

These are minimum requirements. Higher specifications will improve processing throughput and concurrent study capacity.
Memory
64 GB
CPU
16-core processor
GPU
12 GB VRAM (e.g., NVIDIA RTX 4080 or equivalent)
Storage
1 TB SSD (NVMe preferred)
Operating System
Ubuntu 18+, Debian 10+, CentOS 7+, Amazon Linux 2, Windows 11 Pro
Power Supply
800W recommended (typical draw ~675W: GPU 450W, CPU 105W, motherboard 80W, fans/drives ~40W)
Network
Gigabit Ethernet. DICOM routing access to ultrasound devices and PACS.

Security & Compliance

Us2.ai maintains the certifications and compliance standards required for enterprise healthcare deployment.

GDPR & HIPAA

Full compliance with global patient data privacy regulations. BAA available for US customers.

ISO 27001

Certified information security management system. Annual audits and continuous monitoring.

SOC 2 Type II

Independently audited security controls for availability, processing integrity, and confidentiality.

MDSAP

Medical Device Single Audit Program certification for regulatory compliance across multiple markets.

Data Encryption

AES-256 encryption at rest. TLS 1.2+ in transit. Optional customer-managed encryption keys.

Data Retention

Configurable retention policies. Data can be purged on demand. No data is used for model training without explicit consent.

Access Control

Role-based access control. Audit logging for all system access. SSO integration available.

Cybersecurity Review Scope

Common questions from IT security reviews — answered up front.

Does Us2.ai store patient data?
Configurable. Cloud deployments can process in real-time without persistent storage. On-premise retains data within your facility.
What network ports are required?
Standard DICOM port (104/11112) for study routing. HTTPS (443) for cloud connectivity and management. No inbound internet access required for on-premise.
Is a VPN required?
Not for cloud. On-premise operates entirely within your LAN. Optional VPN for remote management access.
How are software updates handled?
Cloud: automatic, zero-downtime updates. On-premise: scheduled update packages deployed by your IT team or Us2.ai support.
What happens if Us2.ai goes down?
Your existing workflow continues unaffected. Us2.ai operates as an additive layer — if the service is unavailable, studies route to your normal reporting path.
Can we run a pilot without production risk?
Yes. Parallel processing mode routes studies to Us2.ai while your existing workflow continues. Zero impact to production until you choose to switch.

Ready to Start Your Evaluation?

Our team can provide a detailed technical architecture review, compliance documentation, and deployment plan specific to your environment.

Contact Sales Engineering