The instrument that records the whole. An observational holon: it observes the system, observes its own apparatus, and observes itself observing, each version of that apparatus immutable, the sequence of versions shifting over time. Holonograph is a patent-pending product of Precision Innovations LLC.
Holonograph complements any evaluation system by adding attribution it can't get on its own. And because it ships as a binary rather than SaaS, it reaches the operators that server-based products structurally rule out:
- regulated industries
- air-gapped and sovereign deployments
- mid-market teams without a platform-engineering org
Holonograph easily accommodates this cohort by nature of its architecture: data sovereignty and methodology depth in a single artifact you deploy and own.
Core capabilities
Mediating-gateway captureobserve by position, never by instrumentation in your code.
Multiplex routingrun one call against several vendors at once and compare them head-to-head on speed, price, and accuracy.
Four-source drift decompositionsubstrate, light source, lens, and noise, with stated-confidence attribution.
Lens versioningthe evaluation apparatus made an explicit, attributable thing.
Closed-loop curationcaptured overrides and approvals cluster into drafted skills, lessons, and fixtures, human-gated before they ship.
Fixture draftingturn observed behavior into the tests that catch the next regression.
Recursive observabilitythe drafter is observed by the lens it improves; getting better is itself a measured event.
Eval-mode-on-productionarchitecturally-enforced run-mode discrimination against production infrastructure.
Vendor-agnostic continuitysurvives model changes and silent reroutes behind an unchanged name.
Variance quantification & isolationrecover true substrate variance by subtracting independently-determined instrument and lens components.
Self-hosted, signed & notarizedyour data never leaves your infrastructure.
What Holonograph is not
Holonograph is frequently confused with adjacent tools. It is none of these:
Not concept-drift detectionthat field assumes a deterministic model and a measurable shift in input data. Holonograph addresses agentic systems on non-deterministic foundation LLMs where the model itself can be silently re-routed by the vendor.
Not feature attributionSHAP, LIME, and integrated gradients attribute a model's output to its input features. Holonograph attributes pass-rate changes over time to one of four mutually exclusive sources of change.
Not an MLOps dashboardnot a tracing system, not a model registry, not a metrics platform. Holonograph composes on top of OpenTelemetry; the novel layer is the versioned lens and the attribution it enables.
Not LLM-as-judgeconventional LLM-as-judge treats the judge as a fixed measuring stick. Holonograph rejects that — it treats the judge itself as a first-class versioned instrument whose drift is independently attributable.
Not a fine-tuning pipelinethe Curation Loop produces operator-curated artifacts (skills, lessons, fixtures, code fixes) that ship through a human-approval gate. It does not adjust model weights.
Not SaaSa signed, notarized native binary that runs as a localhost daemon. Data never leaves the operator's infrastructure.
Observation by position
FIG. 1 — Mediating apparatus: the lens between agent and model.Holonograph sits as a bidirectional mediating gateway between your agents and the models they call, capturing every interaction at the wire-format boundary. Because it owns that boundary, it can run a single call against several models at once (multiplex routing) and compare them head-to-head on speed, price, and accuracy, so you switch vendors on evidence without touching your agents.
An OpenTelemetry sidecar covers anything that doesn't pass through the lens, so the result is observational completeness: every call is either graded or captured, and nothing the agent does is structurally invisible to the apparatus.
No code embedded in the system under observation
No part of Holonograph lives in your agent's code. There is zero evaluation logic embedded in the system under observation, a property of architectural position, not of instrumentation you must install, maintain, and reconcile. That separation is what lets Holonograph run evaluation discipline against production traffic itself, mapping fixtures and gates 1:1 against production responses.
The four sources of drift
Substrate driftoperator-controlled changes to the agentic system.
Light-source driftvendor-controlled evolution of the model, including silent re-routing behind an unchanged model name.
Lens driftoperator-curated evolution of the evaluation apparatus itself.
Stochastic noisethe model's irreducible non-determinism.
Existing practice collapses the first two into "the system regressed," ignores the third entirely, and treats the fourth as either invisible or all-encompassing. Holonograph captures sufficient versioned state across all four at every evaluation event, so any observed change can be attributed to one source with stated confidence. An artifactual gain produced by a change to the apparatus can be told apart from a genuine improvement in the system.
This matters most where it's hardest to see: the vendor of a model cannot honestly grade its own model's drift. The auditor cannot be the auditee. Holonograph is the independent measurement layer that model vendors structurally cannot provide.
The lens as a first-class, versioned instrument
The Lens Architecture treats the evaluation apparatus as a first-class, versioned, independently attributable thing. The lens (the operator-built evaluation surface, with its fixtures, baselines, cohort scheme, and surface contracts) is immutable within each version and replaced rather than mutated.
Because lens changes are discrete, operator-curated, and versioned, the variance contributed by the apparatus over any window becomes a known quantity rather than an unaccounted-for confounder.
Holonograph is the implementation; the Lens Architecture is the framework
It composes on top of OpenTelemetry and is closed under multi-agent composition: handoffs between agents become substrate references rather than new evaluation boundaries. The four-source decomposition holds whether you are observing a single agent or a cooperating swarm.
The Curation Loop
A system's interactions contain the information that should make it better. In conventional practice that information is noted and forgotten, or it demands slow, vendor-dependent fine-tuning. Holonograph closes the loop instead. Every consequential event is captured as first-class ground truth: not only failures, but human overrides (a trainer edits a draft before sending) and human approvals (a trainer reads a draft and sends it as-is). Overrides encode what good looks like right next to what the system produced; approvals are positive signal, not silence.
Similar events cluster, and the drafter (an LLM executed from within Holonograph) proposes a concrete corrective artifact for each cluster: a new skill, a lesson, a fixture, or a code fix. It drafts fixtures too, turning observed behavior into the very tests that will catch the next regression.
Nothing ships unreviewed: every draft lands in an approval gate where a human accepts, edits, or rejects it. Once published, the artifact becomes substrate: versioned, captured, and attributable in future drift analysis like any other change.
The loop is recursive by construction. The drafter's own model call is observed by the same lens it is improving, so the act of getting better is itself a measured, attributable event. Holonograph watches the system, watches the apparatus, and watches itself watching: the instrument that records the whole, applied to its own improvement.
What a pilot looks like.
A pilot is a scoped engagement: deploy Holonograph alongside an existing production agent, instrument the first lens, and run four-source attribution on real traffic. We do the integration and the methodology lift together. Outputs are yours — the binary, the lens versions, the captured substrate. Nothing goes back to a vendor cloud.
What we ask of design partners.
Honest feedback on the methodology and the operator surface, and a willingness to publish at least one finding (anonymised at your call) so the field can compound on it. We sign a mutual NDA, you sign nothing exclusive.
![]()