Ongoing operation of a deployed XyDromatics environment. 24×7
monitoring, patch and upgrade management, capacity planning,
audit-trail attestation. Defined SLAs with named technical
contacts — not a generic helpdesk queue.
Match the tier to your clinical operations. Most hospitals
end up on M2; ambulatory sites on M1; tertiary referral
centers and large health systems on M3. Tiers are per-site —
a multi-site customer can run different sites at different
tiers based on each site’s clinical profile.
Tier M1
Business Hours
Mon–Fri, 8am–6pm Eastern. After-hours best-effort via on-call.
Severity
Response
Resolution target
Sev 1 — clinical-impacting outage
15 minutes (business hours); 1 hour (after hours)
4-hour target
Sev 2 — degraded performance
1 hour (business hours)
8-hour target
Sev 3 — non-impacting issue
4 hours (business hours)
5-business-day target
Sev 4 — request / question
1 business day
Engagement-dependent
Best fit:
Imaging centers and ambulatory sites without 24x7 clinical operations.
Tier M2
24x7 Standard
24x7x365 monitoring. After-hours response covers Sev 1 and Sev 2.
Severity
Response
Resolution target
Sev 1 — clinical-impacting outage
15 minutes, 24x7
4-hour target
Sev 2 — degraded performance
1 hour, 24x7
8-hour target
Sev 3 — non-impacting issue
4 hours (business hours next-day)
5-business-day target
Sev 4 — request / question
1 business day
Engagement-dependent
Best fit:
Hospitals and health systems with overnight imaging operations (ED, ICU, after-hours reads).
Tier M3
24x7 Critical
24x7x365 monitoring + on-call engineer + named secondary on standby. Active synthetic monitoring against your environment with proactive alerting.
Severity
Response
Resolution target
Sev 1 — clinical-impacting outage
5 minutes, 24x7
2-hour target
Sev 2 — degraded performance
30 minutes, 24x7
4-hour target
Sev 3 — non-impacting issue
1 hour, 24x7
3-business-day target
Sev 4 — request / question
4 business hours
Engagement-dependent
Best fit:
Tertiary referral centers, trauma centers, and large multi-site health systems where any imaging downtime translates to clinical impact within minutes.
Response times are clock-on-receipt, not ticket-creation. Resolution
targets are best-effort within scope; complex cross-system
incidents may require coordinated work with your team or
third-party vendors. Service-credit schedule is in the Master
Services Agreement.
Coverage
What’s included, what’s not.
Explicit about both. The clearer the boundary, the fewer
friction points during a real incident — when the question
"is this our problem or yours?" needs to be answered in
minutes, not relitigated across email.
In scope
Operational monitoring
24x7 health monitoring of XyDromatics services
DICOM AE-Title reachability checks
Storage utilization + capacity forecasting
Database health (PostgreSQL / SQL Server / Oracle catalog backends)
Workstation imaging or radiologist workstation support
Custom report development not part of the operational baseline
Major version upgrades that require contractual amendment (typically annual cadence)
Disaster recovery activation — separate DR engagement, scoped per customer
Items above are typically owned by your team, your
infrastructure provider, or another vendor. We’ll
coordinate across boundaries during an incident, but
we’re not the responsible party.
Operational rhythm
A predictable cadence, not surprise activity.
Managed Services runs on a published schedule so you know what
to expect from us and when.
Cadence
Activity
Continuous
Health monitoring + alerting + auto-remediation where defined
Daily
On-call shift handoff with active-incident review
Weekly
Operational status email to your team
Monthly
Operations report + 30-minute review call
Quarterly
Quarterly Business Review (capacity, incidents, roadmap)
SynthGateway — the secure tier underneath managed services.
Every promise on this page — proactive monitoring, 5-minute
Sev 1 response, recorded engineer-initiated remote sessions —
relies on us being able to see + reach your environment fast,
on your terms. SynthGateway is the purpose-built communication
tier that makes that possible.
It replaces the messy alternative — ad-hoc VPNs, screen-share
tools, customer-side firewall openings, vendor-controlled
remote-desktop products. Outbound-only mTLS WebSocket from
each deployed XyDromatics / Synthology product to our gateway
server, two channels, both default-off, both customer-
authorized per signed agreement (DOC-2026-329).
Channel 1 — health_telemetry
Periodic structured health data — system health, disk space,
queue depth, error/log telemetry (PHI-scrubbed at the agent).
Default-off; customer enables when signing managed-services
agreement.
System health checks (service state, daemon liveness)
Revocation takes effect within one heartbeat (≤ 5 min)
Customer-side guarantees.
Both channels are default-off; you authorize
each one explicitly per signed agreement
(DOC-2026-329, Customer Authorization). Either channel can be
revoked from your side at any time — revocation propagates
fleet-wide within one heartbeat. PHI never crosses the gateway
(two-layer scrubber at the agent strips it before transmission).
Full controlled-document set:
DOC-2026-326 (Service Procedure),
DOC-2026-327 (Cybersecurity Threat Model Addendum),
DOC-2026-328 (Technical Specification),
DOC-2026-331 (Engineer Console UX),
DOC-2026-332 (Scaffold Reference). Available under NDA.
Beyond the Synthology fleet
SynthPulse — monitoring for the products we didn’t build.
SynthGateway watches the XyDromatics / Synthology fleet from the
inside — every product reports its own health. SynthPulse extends
that same monitoring to the third-party / other-vendor
products running alongside ours: a foreign PACS, an
interface engine, another vendor’s archive.
It’s a headless, outbound-only agent that watches a product
from the outside — black-box liveness — and reports to
SynthGateway, so those systems surface in the same engineer
console and fleet dashboards as the Synthology products. No code
changes to the target; nothing inbound to open.
Black-box liveness probes
Pluggable checks that run on a schedule and post their results —
nothing to embed in the monitored product.
http — reachability, HTTP status code, and TLS certificate expiry
tcp — raw connect on any port (HL7/MLLP, a database, a DICOM listener at the transport level)
dicom-echo — a real DICOM C-ECHO: proves the peer accepts a DICOM association and answers Success (a port-open check can’t)
Pluggable framework — HL7-MLLP, SNMP, and SSH probes as fast-follows
Deploys anywhere, reports to your gateway
One self-contained binary; the gateway is the control plane.
Headless — a Windows service, a Linux systemd daemon, or a one-shot cron job
Outbound-only — no inbound firewall openings on the monitored network
Token-licensed — entitlement is a scoped, revocable gateway ingest token; no license file, no per-binary key
PHI-free by design — numeric metrics with charset-validated names; each target appears as an external agent alongside the fleet
Push when you can, probe when you can’t.
A third-party product that can report on itself POSTs its own KPIs
straight to the SynthGateway ingest API; one that can’t gets
watched by SynthPulse from the outside. Both land in the same
monitoring, alongside the Synthology fleet. General-purpose
monitoring infrastructure — not a medical device, PHI-free by
design. Selection guidance: DOC-2026-485 (Ingest API vs SynthPulse).
Available under NDA.
Managed research de-identification
Send us PHI over a secure channel; get research-safe data back.
For institutions that want a de-identified research data set
without standing up and operating the pipeline themselves,
Synthology hosts the
De-Identification Engine
and VNA Research in our cloud. You push studies and reports to us
over a mutually-authenticated, encrypted channel; we de-identify
at the trust boundary and hand back a Research-Use-Only data set.
You keep the PHI & the key
Raw PHI crosses only inside a fail-closed, mutually-authenticated
channel (DICOM-TLS C-STORE, MLLP-over-TLS, or DICOMweb STOW-RS).
The re-identification key lives only on the engine — nothing
re-identifiable is ever stored in our cloud. Only de-identified
output persists.
Expert-Determination de-id
De-identification follows the HIPAA Expert-Determination method
(§164.514(b)(1)): keyed pseudonymization, interval-preserving
date shift, free-text and burned-in-pixel scrubbing, and coded-
concept harvesting — all operated by Synthology under a BAA.
Cohorts under two-person control
Coded-concept cohorts export only under a current Expert-
Determination attestation and an enforced k-anonymity floor —
the same two-person control whether you self-host or we operate
it for you.
Prefer to run it inside your own network? The De-Identification
Engine is also a customer-deployable product — see the
product page.
Non-device software under FD&C Act §520(o)(1)(D); de-identified
output is Research-Use-Only.
Pricing examples
Annual subscription, scaled to deployment.
Pricing is annual subscription tied to the deployed product
surface area, modality count, and SLA tier. Multi-year commits
(24 / 36 months) carry rate locks and can save 5–15%. Indicative
ranges below; real customers are quoted off scope after a
discovery conversation.
Tertiary referral center, 9-product XyDromatics deployment
Tier M3 (24x7 Critical)
$45,000 – $75,000 / month
Onboarding (4–6 weeks before the SLA clock starts) is a
separately-quoted project. Disaster Recovery activation
(DR plan execution under a real-event scenario) is its own
scoped engagement, not part of the standing subscription.
Frequently asked
The questions we get most.
Why is Managed Services XyDromatics-only?
Two reasons. First, Managed Services covers the operational lifecycle (patches, upgrades, license reconciliation) of the specific products we ship — we have deep expertise on our own codebase and can't commit to the same depth on a competitor's product where we don't see source. Second, the SLA structure depends on direct access to product engineering for sev-1 escalations. For non-XyDromatics environments, the right service is Professional Services staff augmentation.
How does Managed Services interact with our existing IT team?
We're an extension of your team, not a replacement. Your team still owns infrastructure, network, and the broader imaging stack. We own the operational lifecycle of the XyDromatics products specifically. Most customers run a weekly sync where our named contact joins their imaging-IT standup.
What's the contract length?
Standard contract is 12 months with auto-renewal. Multi-year commits (24 / 36 months) get rate locks and can save 5–15%. 90-day exit clause for cause — we don't lock you in if the service is materially degraded.
Do you take over our existing XyDromatics deployment, or only ones you implemented?
Either. Most Managed Services customers come through our Implementation Services tier, but we also onboard environments deployed by other implementation partners. Onboarding engagement is a 4–6 week scoped project that produces the environment runbook before the SLA clock starts.
What happens during an SLA miss?
Service credits per the SLA terms. Repeated misses inside a rolling 12-month window trigger an escalation to your executive sponsor and an internal RCA on our side. The full SLA terms with response-time definitions and credit schedule are in every Master Services Agreement.
Where does your team work from?
Synthology team members are US-based, primarily East Coast and Mountain time zones. On-call coverage is provided by named Synthology engineers — not by an offshore NOC. The higher SLA tiers carry higher cost partly because we don't subcontract on-call to a generic helpdesk.
How do you handle data access for managed environments?
Least-privilege: Synthology engineers have read-only access by default; write access is granted per-incident with audit logging. Sensitive PHI access is gated by a named-engineer allowlist that you review quarterly. Full access policies are in the Managed Services SOW addendum.
How do you actually connect to our environment? Do we have to open firewall ports or run a VPN?
Neither. SynthGateway is the secure communication tier built specifically for this — outbound-only mTLS WebSocket from each deployed Synthology / XyDromatics product to our gateway server. No inbound firewall openings on your side. Two channels, both default-off, both customer-authorized: health_telemetry (PHI-scrubbed structured health data; enables our 24x7 monitoring) and remote_support (engineer-initiated SSH/HTTPS/RDP sessions, recorded end-to-end; enables fast Sev-1 response). You can revoke either channel from your side at any time; revocation propagates within one heartbeat (≤5 min). See the "How we connect" section above for the full architecture, or ask for the controlled-document set (DOC-2026-326..332) under NDA.
Can we transition from Managed back to self-managed later?
Yes. The runbook produced during onboarding is yours, and we provide a 60–90 day knowledge-transfer engagement at exit. No proprietary tooling locks you in — operations run against standard XyDromatics interfaces that any qualified team can operate.
Talk through a Managed engagement.
Tell us about your deployed environment and clinical
operations profile. We’ll come back within one business
day with a proposed SLA tier and the named technical contact
who would own the relationship.