Services / S3 · Sustain
Continuous AI Monitoring.
Bias isn’t a static property. It shifts over time through data drift, user behavior, and feedback loops — we call it bias creep. Test once and stop watching, and you’re trusting a false sense of safety. Continuous Monitoring meets the post-market monitoring duty under Art. 72 and gives you the evidence base for quarterly board updates and ESG reporting.
Who it’s for
Your AI is live — and has to stay defensible.
Monitoring is the right step when your systems are in production, carry high-risk characteristics or ESG relevance, and you need regular, documented evidence for investors, rating agencies, your board, or EU customers who expect ongoing assurance — not a one-time certificate.
Post-market obligation
Art. 72 requires providers of high-risk systems to monitor performance after deployment. Monitoring operationalizes that duty instead of leaving it as a line in a policy.
Models that change
Retraining, shifting inputs, and feedback loops move behavior over time. What was fair at audit can drift — and you want to know before a customer or regulator does.
Recurring reporting
Boards, investors, and CSRD-relevant stakeholders ask on a cadence. Monitoring produces the quarterly evidence rather than a scramble each reporting cycle.
How it runs
Set up once, then operated as a service.
We build the pipeline on open-source foundations you own, then operate or co-operate it — no vendor lock-in.
Pipeline & thresholds
We instrument the system: drift detection on inputs and outputs, temporal fairness metrics with alert thresholds agreed with you, and a data quality pipeline. We define what “out of bounds” means for your context and who gets alerted when it happens.
What you’ll have
- Drift detection (Alibi Detect / Evidently AI)
- Temporal fairness metrics with alert thresholds
- Data quality pipeline (Great Expectations)
- Escalation paths and accountability matrix
Operate & alert
The monitoring runs continuously against a dashboard built for compliance owners — not just data scientists. When a threshold is crossed, the right people are alerted along the escalation path, with enough context to act.
What you’ll have
- Monitoring dashboard for compliance owners
- Automated threshold alerting
- Incident log with context for each alert
Quarterly compliance review
Every quarter we produce a review report fit for the board, investors, or an EU customer’s ongoing due diligence — what changed, what was caught, what was done. This is the documented evidence the post-market duty calls for.
What you’ll have
- Quarterly compliance review report
- Trend analysis across the period
- Recommendations and remediation tracking
Investment
A setup, then a retainer.
Setup cost scales with the number of systems and pipeline complexity; the monthly retainer reflects the cadence of operation and reporting.
Single system
One production system, instrumented with drift detection, fairness metrics, and a compliance dashboard, plus quarterly reviews.
from €6,000 setup
Portfolio · custom
Multiple systems
A portfolio of production systems under one monitoring program, often as part of a Transformation engagement. Priced after a scoping call.
Custom estimate
Built on open-source foundations you own. Bundle discounts apply within a full engagement (–20%). All prices net of VAT.
AI governance isn’t a project with an end date. It’s a state you keep — and prove you’re keeping.
— Dr. Valentin José Mayr · Founder
Related services
What feeds the monitoring.
S1 · Locate
Readiness Assessment
Identifies which systems carry the post-market duty in the first place.
from €15,000
S2 · Test
AI Ethics Audit
Establishes the fairness baseline that monitoring then watches over time.
from €8,000
S4 · Understand
Workshops
Train your team to read the dashboard and act on alerts confidently.
from €2,000
Get concrete