Pilots & use cases
The instrument is currently being evaluated across different environments, technologies and use cases, around a single question:
Can a single measurement primitive become a shared resource for multiple systems, without modifying their infrastructure?
The pilots presented below show how the NeoMundi signal can be consumed to build agentic control, audit, evidence, governance, insurance or behavioral tracking functions.
Each activation is presented by use case, targeted value and stage of progress — validated pilot, in progress, MVP or available solution.
One shared signal. Different environments. Value built downstream.
Observe · Detect · Multi-agent · Evaluate · Control · Prove · Govern · Insure · Optimize
We produce the signal. You build the value around it.
Our full documentation →Pilot technologies and environments
01 · Insurance and risk evidence
In progress
NeoMundi measures → the methodology cross-references rules, permissions, traces and severity → an auditable risk file can be built
What the pilot evaluates: the ability to document whether an AI agent acted in line with its permissions and declarations, in order to build a usable evidence base for insurance analysis.
Technical evidence — published after validation
02 · Agentic control and routing
Validated pilot · MVP
NeoMundi measures → REVAS consumes the signal according to defined policies → continue, verify, regenerate, switch models, escalate or stop
What the pilot demonstrates: an independent orchestration technology can consume the NeoMundi signal to adapt how an execution is processed.
Technical evidence — published after validation
03 · Policy and enforcement
Validated pilot
NeoMundi measures → OASSE/FAM applies deterministic rules → Gatekeeper allows, holds, redirects or blocks
What the pilot demonstrates: the NeoMundi signal can feed an independent policy layer, then an enforcement mechanism, without itself becoming the decision-making authority.
Technical evidence — published after validation
04 · Controlled AI agent audit
Validated pilot
NeoMundi measures → the test purchase cross-checks the agent’s actions, log and receipt → discrepancies can be documented and reviewed
What the pilot demonstrates: behavioral measurement and receipt compliance are two distinct, complementary variables.
Synthetic reference environment.
Technical evidence — published after validation
05 · Evidence reconstruction
Validated pilot
NeoMundi measures → FactNotebook ingests the signal and its provenance → evidence and governance objects are reconstructed
What the pilot demonstrates: an independent technology can integrate a NeoMundi measurement as an evidence item without altering its provenance.
Technical evidence — published after validation
06 · Parallel observability
Validated pilot
The same real execution → NeoMundi and DriftShield/HIC observe it independently → observations remain distinct and cross-referenceable
What the pilot demonstrates: two infrastructures were able to observe the same real execution without sharing the raw response and without modifying the existing X/Y/Z authority.
Limitation: minimal validation on a single real execution.
Technical evidence — published after validation
07 · Governance through verifiable evidence
In progress
NeoMundi measures → SEILX captures, preserves and anchors the artifacts → the retained evidence can be checked later
What the pilot evaluates: the verifiability of the integrity, provenance and time-anchoring of evidence associated with an observed event.
Technical evidence — published after validation
08 · Admissibility and execution continuity
Architecture documented · Validation in progress
NeoMundi observes → RGE-01 and REB-01 assess admissibility and govern activation → RiCo-01 handles integrity and continuity after activation
What the use case establishes: measurement, admissibility, execution authority and continuity can remain separate architectural responsibilities.
Full machine-to-machine validation not yet available.
Technical evidence — published after validation
From an intuition, on September 27, 2025, to our first pilots and partners.
Are you developing a technology, deploying an AI system, or exploring an audit, evidence, governance, risk management or insurance need?
Propose a concrete case to us. Selected projects go through a targeted, limited and documented pilot to evaluate the signal’s interoperability and the operational value built around it.
Your technology. Your use case. Your decisions. Our measurement signal.
