Optronix Optronix
←→ navigate F fullscreen contents

Managed
operations.

Data centre and network operations for GPU and high-performance colocation estates.
Service pack    Commercial in ConfidenceRev 1.0    Prepared by Anthony Hobley, 2026
→  to begin    F for fullscreen
Contents

What we'll cover.

    The proposition

    As much of the floor and the fabric as you want to hand over.

    One hall, two services Goods in and spares Overhead containment Busbar, PDUs and metering Environmental sensors Racks, hardware and the asset record Raised floor Spine Spine Uplinks to the edge Leaf switches

    Not remote hands Remote hands is a colocation commodity, sold by the fifteen minutes and measured by the ticket. This is an operating team that holds your estate, whether or not anything has broken.
    Who we are

    Where our engineers have worked.

    USAUKEuropeMiddle EastAustralia
    Delivering today

    Not a proposal. A service that is already running.

    Scheduled data centre operations are live today in five cities across two continents, at a range of service tiers. These are the sites we run now, not the limit of where we can run.

    01 02 03 04 05
    Reach The tiers are not tied to these cities. We mobilise to wherever the estate is: EMEA, North America and further afield, across any number of sites.
    Service one

    Data centre
    operations.

    The floor: racks, cabling, power, hardware and logistics, and the record of all of it.

    Data centre operations

    Five tiers, rising on presence and depth.

    01Copper

    On-Call

    We attend on request, with the runbook and the asset register already held and maintained between visits.

    On siteOn request
    Response48 hr / NBD
    02Bronze

    Scheduled

    Fixed attendance days per site, working a managed ticket queue, with reactive cover in between.

    On siteSet days per week
    Response48 hr
    03Silver

    Managed

    Two or more days a week on site, plus 24/7 monitoring, and we own the incident through to closure.

    On site2+ days per week
    Response8 hr
    04Gold

    Resident

    Technicians on site every day across an extended two-shift pattern, on call overnight.

    On siteDaily, two shifts
    Response30 min
    05Platinum

    Embedded

    A dedicated crew on site around the clock, with spares held on site and a committed availability target.

    On site24/7/365
    Response15 min
    Presence

    Who is on site, and when.

    On site On call 24/7 monitoring
    Time to hands on

    Attendance patterns are illustrative. Days, hours and shift times are set per site. The fifteen-minute detection and ticketing promise applies at every tier.

    Data centre operations

    What each tier includes.

    01On-Call02Scheduled03Managed04Resident05Embedded
    Installs, moves, adds and changes●●●●●
    Ticketing portal, managed queue and service reporting●●●●●
    Goods receipt, shipping and RMA logistics○●●●●
    Asset register ownership and audit○●●●●
    Photographic evidence on every change○●●●●
    24/7 infrastructure monitoring○○●●●
    Incident ownership through to closure○○●●●
    OEM and warranty escalation management○○●●●
    Named service delivery manager and governance reporting○○●●●
    Live DCIM portal for power, thermal and capacity○○○●●
    On-site spares pool held and managed○○○●●
    Firmware, BIOS and driver lifecycle○○○●●
    Environmental sensor mesh integrated into DCIM○○○●●
    Fabric health across InfiniBand and NVLink○○○○●
    Node validation and return to service○○○○●
    Full estate telemetry and two-way ITSM ticket sync○○○○●
    Consignment spares, funded and stocked○○○○●
    Committed fleet availability target○○○○●
    ●Included○Available as an option
    These are the off-the-shelf tiers. Every one of them is a starting point for a service built around your estate.
    Service two

    Network
    operations.

    The fabric: monitoring, configuration, patching and change, on whatever is already in the racks.

    Network operations

    The same ladder, drawn on the fabric.

    01Copper

    Monitor

    We watch the fabric around the clock and raise the ticket the moment something breaks. Your team keeps the hands.

    Scheduled visitsNone
    ResponseChargeable
    02Bronze

    Manage

    We hold the configuration, run the patching and execute standard changes on a managed queue.

    Scheduled visitsNone
    ResponseSubject to availability
    03Silver

    Managed Response

    Full change ownership and 24/7 remote engineering that resolves rather than escalates, plus a monthly walk of the estate.

    Scheduled visitsMonthly
    Response48 hr, included
    04Gold

    Assured

    Proactive capacity, performance and vulnerability management, a named service manager, and a visit every fortnight.

    Scheduled visitsFortnightly
    Response8 hr, 24/7
    05Platinum

    Critical

    Weekly attendance, spares on site, quarterly failover testing and a committed availability target.

    Scheduled visitsWeekly
    Response4 hr, 24/7
    Network operations

    What each tier includes.

    01Monitor02Manage03Managed Response04Assured05Critical
    24/7 monitoring of device, interface and control-plane health●●●●●
    P1 outage notified and ticketed within 15 minutes●●●●●
    Fault triage and diagnosis●●●●●
    Ticketing portal, managed queue and service reporting●●●●●
    Configuration backup, restore and drift detection○●●●●
    Vendor TAC case ownership and escalation○●●●●
    Network OS patch management in scheduled windows○●●●●
    Standard change execution○●●●●
    Fabric controller operation and assurance○●●●●
    Normal change, RFC preparation and CAB representation○○●●●
    24/7 remote engineering support○○●●●
    Incident ownership through to resolution○○●●●
    Vendor advisory and vulnerability exposure tracking○○○●●
    Proactive capacity and performance management○○○●●
    Named service manager and quarterly service review○○○●●
    Live fabric telemetry streamed into your view○○○●●
    Two-way ticket sync into your ITSM over API○○○○●
    Quarterly failover and resilience testing○○○○●
    Consignment spares held on site○○○○●
    Committed fabric availability target, where the tier supports it○○○○●
    ●Included○Available as an option
    These are the off-the-shelf tiers. Every one of them is a starting point for a service built around your estate.
    Service levels

    What fifteen minutes actually means.

    Detected, notified and ticketed 15 min, every tier Time to fix Not a response target Event Longest response target 15 min Hands on site

    Time to hands on, not time to fix. A response target that quietly pretends to be a fix target is a number nobody can hold, and everybody finds that out on the first bad night.

    A response target is worthless without spares. Fifteen minutes to a rack you cannot fix is fifteen minutes. That is why on-site spares are inside the top tier rather than sold beside it.

    We never run one-up. Lone working is banned on live electrical and liquid-cooled plant, and a single technician on shift means the second concurrent incident breaches by definition.

    Every tier, one promise

    A P1 outage is detected, notified and ticketed within fifteen minutes. That applies at Copper exactly as it does at Platinum.

    The tier changes what happens next, not how quickly you hear about it.

    15 min, every tierDetected and ticketed
    48 hr → 15 minOn-site response, by tier
    From tier 03Incident ownership
    Taking both

    One team, one queue,
    no gap in between.

    One queue

    An alert on a switch and a body in the hall are the same ticket, on the same SLA clock, in the same portal.

    No handoff

    Nobody gets to say the fault is on the other side of the patch panel. One contract, one service manager, one number.

    One mobilisation

    One team badged, inducted and cleared for the site once, covering the floor and the fabric from the same shift.

    Not a straitjacket The tiers are a starting point. Any capability can be added to any tier, and the mix can differ site by site. Taken within the tier it is better value than bolted on.
    The portal

    One place to see all of it.

    Every tier includes the portal: raise and track tickets, interrogate the asset register and pull service reports on demand. From Gold we instrument the estate itself; at Platinum you also get two-way ticket sync into your own ITSM over API, so an event on the floor opens, updates and closes in your system without anyone typing.

    What we instrument

    On the floor
    • Branch circuit power
    • Inlet and exhaust thermals
    • Differential pressure
    • Leak detection
    • Door and cabinet state
    • Rack-level telemetry
    • GPU and fabric health
    • Capacity and headroom

    What we monitor

    On the fabric
    • Device and supervisor health
    • Interface state and errors
    • Control-plane adjacency
    • Optic light levels
    • Configuration drift
    • OS version and advisory exposure
    • Environmental and power draw
    • Capacity and port headroom
    The tooling saasyn The portal, the DCIM and the monitoring are our own, built with our software partner. If you would rather keep what you already run, we will operate yours, migrate you onto ours, or integrate the two so the whole estate reports into one place.
    The portal

    Dashboard

    Uptime, power, PUE, thermals and open work across the estate, at a glance. Live in the browser, not a screenshot.

    portal.optronix.co.uk/dashboard
    The portal

    Service desk

    Every ticket, its SLA clock and the full activity trail, including what the engineer did on the floor, and when.

    portal.optronix.co.uk/tickets
    The portal

    DCIM

    Live rack elevations, occupancy, draw and inlet temperature per rack, with the GPU cluster mapped where it actually sits.

    portal.optronix.co.uk/dcim
    Day one

    Deploy and commission.

    From choosing the hall to handing over a running estate. Priced as a project rather than a subscription, and the on-ramp to every tier.

    Step 04
    Hall selected Goods in Containment, power and data As-built record
    01

    On the fabric The network equivalent runs the same shape: HLD validation, low-level design, fabric build and controller configuration, tested against the LLD, as-built documentation and formal handover.
    Already running

    Or we take on
    what is already on the floor.

    Not every estate starts empty. Where the floor is already live, we do not rebuild it. We learn it, record it and take it on.

    01

    Survey and audit

    We walk the floor and check what is actually installed against what the records claim, photographed as we go.

    02

    Record and baseline

    Every asset, circuit and patch into the register, with power and thermal baselined so drift becomes measurable.

    03

    Integrate

    Monitoring and DCIM connected and tickets flowing, with your existing tooling either operated as it stands or migrated.

    04

    Take control

    We assume the runbook, the escalation path and the SLA, and the estate is in steady-state operations.

    No rip and replace The estate keeps running throughout. The only thing that changes is who is holding it, and from that point it is the same service, the same portal and the same tier ladder as a floor we built ourselves.
    Add-ons

    Available at any tier.

    Nine modules that sit alongside the tier tables. Every one of them can be added to any tier, and the mix can differ site by site.

    01Colocation and landlord managementWe hold the site relationship with the provider and the landlord.
    02Single point of OEM contactOne route into every vendor, with the case owned by us.
    03Capacity and power reportingPower, thermal and capacity headroom reported on a cycle.
    04Licence and support contract managementRenewal dates, entitlements and support cover tracked.
    05Consignment sparesFunded, stocked and held on site.
    06Out-of-hours change windowsStandard and normal change executed outside your hours.
    07Decommission and refreshKit taken out, replaced, and the asset record updated.
    08Secure media destructionCertified erasure or destruction, with the certificate.
    09Quarterly failover and resilience testingFailover and resilience proved to a quarterly schedule.
    Assurance, every tier Photographic change evidence Full access chain of custody Background-checked named personnel One pane of glass
    Our work

    The floors we look after.

    A finished hall aisle
    01A finished hall, aisle by aisle
    Patch fields, mid-deployment
    02Patch fields, mid-deployment
    Server cabling, dressed to the rack
    03Server cabling, dressed to the rack

    Our engineers, our projects. Not stock photography.

    Why Optronix

    The team that builds them is the team that runs them.

    We built the estate

    Design, fit-out, cabling and commissioning are the same business. Operations inherits a facility we already know, rather than one we are reading about.

    Named people, not a pool

    The engineers on your floor are background-checked, badged and yours. You get the same faces, and they learn the estate.

    Evidence, not assurances

    Photographic evidence on every change, a maintained asset register, and reporting you can hand to your own auditors.

    Deep dives

    More detail, deck by deck.

    Before day oneData Centre Design & BuildFrom the hall to a commissioned estate, built to a drawing.Open deck → The physical layerStructured CablingCopper and fibre designed, installed, tested and certified.Open deck → On the floorRack, Stack & PatchRacked to drawing, patched to schedule, recorded as built.Open deck → Moving an estateData Centre MigrationsAn existing floor lifted, moved and brought back into service.Open deck → Power and coolingM&E EngineeringElectrical and mechanical works behind the whitespace.Open deck → Before it shipsStaging & ConfigurationKit built, configured and tested before it reaches site.Open deck → End of lifeITAD & WEEESecure decommission, certified erasure and disposal.Open deck → Beyond the data centreWorkplace TechnologyEvery piece of technology in the office, duct to desk.Open deck → ProofCase StudiesProjects with the results, and what the clients said.Open deck →
    Next step

    Tell us the site.
    We'll scope the tier.

    Most estates begin at Scheduled or Managed and step up as the deployment grows. Give us the footprint and the hours that matter, and we will come back with the tier that fits.

    Anthony Hobley
    Technical Director
    +44 (0) 7450 337795
    anthony.hobley@optronix.co.uk     Email Anthony →     Matt Stockill
    Sales Manager
    +44 (0) 7768 360149
    matt.stockill@optronix.co.uk     Email Matt →
    optronix.co.ukEvery service, across the data centre lifecycle Company Overview DeckWho we are, and everything else we do LinkedInWhat we are working on, as it happens

    Data Centre Lifecycle Experts