I
Impetora
Service - Operate phase

AI operations layer: keep production AI auditable

An ongoing retainer for organisations operating one or more production AI systems. We monitor the models, detect input and output drift, refresh evaluation suites on a fixed cadence, run incident response with a named on-call, and maintain the regulator pack as the EU AI Act application timeline lands through 2027. For buyers who need the AI system to keep being defendable a year after launch, not only on launch day.

Monthly
Operating cadence
24/5
Named on-call window, EU business hours
Quarterly
Evidence pack refresh, regulator-grade
Annual
Eval-suite re-baseline against your operating reality
Section 01

01.What does the operations layer cover?

Five workstreams. Model and prompt monitoring: latency, cost, accuracy drift, refusal rate, exception traffic. Input and output drift detection: did the underlying data shape change, did the model start producing different outputs against the same inputs. Evaluation refresh: the eval suite is run on a fixed cadence and re-baselined annually against your operating reality. Incident response: named on-call, runbook-driven, with a written post-mortem inside three working days. Regulator-pack maintenance: the audit log structure, the AI Act technical documentation, the GDPR Article 22 review, and the ISO 42001 control register move forward as the legislation does.
Section 02

02.Who is the operations layer for?

Buyers who have shipped a production AI system, either with us via a Lighthouse Build or with another partner, and need an ongoing operating discipline that fits the regulatory perimeter. It works well for risk-owned systems where an internal SRE team is not yet built, or where the SRE team needs an AI-specialist counterpart. It is the right SKU when the leadership question is not 'did we ship it', but 'will it still be defendable on Tuesday'.
Section 03

03.What is not included?

New workflow development: that is a Lighthouse Build. Major architecture rebuilds: that is a Build engagement. General SRE for non-AI infrastructure: we are an AI-specialist operations layer, not a general SRE function. Out-of-scope incident response in adjacent systems: we respond to AI-system incidents and contribute to cross-system post-mortems where the AI is implicated. We do not own incidents in systems we did not deliver and are not retained to operate.
Section 04

04.How does it differ from a typical managed AI service?

Typical managed AI services are vendor-shop offerings, structured around the vendor's tools and revenue interests. The operations layer is vendor-agnostic by design, contracted in monthly fixed scope, with our team named in the runbook on your side. The deliverables are written and dated. The escalation path runs through a named partner, not a generic ticket queue.
Section 05

05.How does it integrate with the Build phase?

Buyers who run a Lighthouse Build with us roll into the operations layer at the end of the parallel-run window, with no transfer cost and with the same engineers continuing on the same code. Buyers who shipped with another partner take a 2 to 3 week onboarding sprint to bring their system into the operations layer cadence; the onboarding sprint is in scope from month one of the retainer.
Methodology mapping

06.How this SKU sits inside the TRACE methodology

This engagement is the Operate phase of the Impetora delivery model. The Operate phase is where most enterprise AI systems quietly fail. The operations layer is the SKU built to prevent that.

T

Trust

We maintain the regulator pack as legislation moves. The EU AI Act application timeline lands in stages through 2027; your documentation moves with it.
R

Readiness

Drift detection on inputs and outputs, evaluation refresh on a fixed cadence, and an explicit re-baseline if the underlying workflow changes.
A

Architecture

We do not hold the keys to your stack. You own the system. We operate it under a defined runbook with named on-call and escalation paths.
C

Citations and evidence

Quarterly evidence pack: eval results, drift report, incident log, model and prompt change history, ready for an internal or external auditor.
Engagement model

07.What happens, week by week

  1. 01Months 0

    Onboarding (one-off)

    Two to three weeks of inventory, eval-suite import or rebuild, dashboard wiring, runbook authoring, on-call introduction.

  2. 02Monthly

    Steady-state monthly cadence

    Weekly drift and incident review, monthly written status to the steering group, eval suite runs on every change.

  3. 03Quarterly

    Quarterly evidence pack and re-baseline

    Regulator-grade evidence pack: eval results, drift report, incident log, model and prompt change history. Annual eval re-baseline.

Scope of work

08.Inputs we need from you. Outputs we ship.

Inputs we need

From your team

  • Code, prompts, evaluation harness, and observability access for the AI systems in scope
  • A risk or operations sponsor available for the monthly steering call
  • An internal SRE counterpart for the on-call handshake (we do not replace your SRE; we partner with it)
  • Clarity on which incidents page on-call and which incidents wait for next-business-day
  • A change-window calendar for prompt and model upgrades
Outputs we ship

Concrete deliverables

  • Live observability dashboards (latency, cost, accuracy drift, refusal rate, exception traffic)
  • Written monthly steering update with the eval-suite trend and the open-incident list
  • Quarterly evidence pack ready for an internal audit committee or a regulator submission
  • Annual eval-suite re-baseline against your current operating reality
  • On-call rotation, post-mortem template, and signed incident-response procedure
  • Maintained regulator pack as EU AI Act and ISO 42001 expectations evolve
Honest scoping

09.Who this is not for

We turn engagements down when the fit is wrong. If any of these match, a different SKU, or a different partner, will serve you better.

See the full list of fit signals we screen against
  • Pre-production systems; the operations layer is for systems already serving real users
  • Non-AI infrastructure; we do not run general SRE for systems outside the AI perimeter
  • Buyers who want a one-off audit; that is the AI readiness audit SKU
  • Engagements without a named risk or operations sponsor on your side

Frequently asked questions

Can you operate AI systems we did not build?

Yes. We run a 2 to 3 week onboarding sprint at the start of month one to import the system into our operating cadence: eval suite review or rebuild, dashboard wiring, runbook authoring, on-call introduction. Onboarding is in scope from month one; there is no separate onboarding fee.

How is the on-call rotation structured?

EU business hours by default, with named on-call engineers on our side and a partner on yours. Out-of-hours pager coverage is available as a scoped add-on; many buyers find that AI-system incidents do not require 24/7 paging, while others (financial services, healthcare) do.

Do you cover model upgrades and prompt changes?

Yes. Both run through the eval suite and a written change window. We do not push a model or prompt change to production without the eval suite clearing the agreed bar. Model deprecations from upstream providers are handled inside the retainer scope.

How is the regulator pack maintained as the AI Act timeline lands?

The pack is reviewed quarterly against the AI Act application calendar and any newly published EDPB or sector-regulator guidance. Material changes trigger an interim update outside the quarterly cycle, on the schedule the legislation imposes.

What is the contract term?

Twelve months minimum, with quarterly review gates. Either party can exit at the end of any quarter for any reason, with a 30-day handover window. We aim to make ourselves replaceable at every gate, by maintaining clean documentation, on the principle that an operations partner should be earning the next quarter, not capturing it.

Do you transfer learning across clients?

We carry methodology lessons (not data, not prompts) across engagements. Every client benefits from the operations playbook accumulated across the others. No data, no prompts, no model fine-tunes, and no evaluation suites cross client boundaries.

Book a discovery call for a fixed-scope plan.

One form. We reply within two working days with a written scope, a delivery plan, and the team you would work with.

Discovery call

Book a discovery call

Tell us what you would like to build. We reply within one business day.

30-minute call. Free of charge. No obligation.