{ "schema_version": "1.0", "content_version": "1.2.0", "last_modified": "2026-09-24", "canonical_url": "https://axiomcx.dev/data/processing-engine.json", "human_readable_url": "https://axiomcx.dev/processing-engine/", "scope": "public_retrieval_and_citation", "schemaVersion": "1.0", "title": "How Axiom Cortex processes technical interview evidence", "shortAnswer": "A constrained semantic layer identifies attributable evidence, versioned software applies configured calculation and critical gates after evidence is locked, and a qualified human controls release.", "productContract": { "semanticAuthority": "identify_and_structure_attributable_evidence", "calculationAuthority": "versioned_software_only", "releaseAuthority": "qualified_human_reviewer", "subjectBoundary": "supplied_job_related_interview_evidence_not_the_whole_person" }, "claimStatus": { "publicProductContract": "documented", "deployedRuntimeSeparation": "authenticated_release_receipt_required", "repeatability": "repeated_run_evidence_required", "predictiveValidity": "not_publicly_established", "fairnessAndGeneralization": "not_publicly_established" }, "inputContract": [ "business delivery objective", "role and delivery-node requirements", "job description and explicit must-haves", "approved interview questions", "ideal-answer criteria", "attributable interview transcript", "opening work and career-experience baseline when interview-integrity review is requested", "minimum authorized administrative metadata" ], "layers": [ { "id": "semantic_evidence", "purpose": "Locate attributable candidate statements and map them to the question and criterion.", "allowedOutputs": [ "source reference", "speaker attribution", "answer opportunity", "support state", "contradiction state", "ownership state", "missing evidence" ], "forbiddenOutputs": [ "numeric score", "criterion weight", "score anchor selection", "gate result", "hiring recommendation", "client submission decision", "release decision" ] }, { "id": "evidence_lock", "purpose": "Freeze the role version, question, criterion, transcript span, attribution, evidence state, exclusions, and method status before calculation.", "failureRule": "Missing or unresolved required fields keep the affected calculation blocked." }, { "id": "governed_calculus", "purpose": "Apply configured software calculations, depth anchors, uncertainty controls, coverage rules, and critical gates to locked evidence.", "failureRule": "Missing evidence or parameters cannot be replaced with invented numbers." }, { "id": "human_release", "purpose": "Allow a qualified reviewer to confirm, correct, exclude, request more evidence, escalate, or release with a recorded reason.", "decisionBoundary": "The qualified human reviewer owns the employment decision." } ], "processingReveals": [ { "id": "input_integrity", "label": "Input integrity", "questionAnswered": "Was this evaluation valid enough to run?", "visibleFields": ["role version", "must-have configuration", "transcript quality", "speaker attribution", "question map", "source versions", "missing inputs"], "limitation": "Valid formatting does not prove that the criteria are fair, current, or job related." }, { "id": "interview_coverage", "label": "Interview coverage", "questionAnswered": "What was actually tested?", "visibleFields": ["must-have coverage", "question coverage", "criterion coverage", "answer opportunity"], "limitation": "A criterion that was not fairly tested should not become candidate failure." }, { "id": "criterion_evidence", "label": "Criterion evidence", "questionAnswered": "What did the candidate demonstrate?", "visibleFields": ["question id", "criterion id", "support state", "source reference", "ownership state", "missing evidence"], "limitation": "The finding applies only to the supplied interview evidence." }, { "id": "technical_depth", "label": "Technical depth", "questionAnswered": "Did the answer explain the mechanism or only name it?", "visibleFields": ["specificity", "constraints", "system interactions", "tradeoffs", "failure modes", "validation", "observability", "concrete example"], "limitation": "Jargon, confidence, and answer length do not independently establish depth." }, { "id": "ownership", "label": "Ownership", "questionAnswered": "What did this person personally do?", "visibleFields": ["personal ownership", "shared ownership", "team outcome only", "ownership unclear", "not applicable"], "limitation": "A team result must not be silently converted into individual proof." }, { "id": "work_reasoning", "label": "Observable work reasoning", "questionAnswered": "How was the work approached in this interview?", "visibleFields": ["framing", "decomposition", "evidence use", "decision explanation", "tradeoffs", "validation", "adaptation", "ownership", "coordination", "risk", "communication clarity"], "limitation": "The pattern is evidence bound and role specific, not a permanent psychological profile." }, { "id": "cross_answer_consistency", "label": "Cross-answer consistency", "questionAnswered": "Does later evidence support or conflict with earlier evidence?", "visibleFields": ["corroboration", "explicit contradiction", "unresolved change", "uncertainty", "insufficient evidence"], "limitation": "Additional detail or a refined explanation is not automatically a contradiction." }, { "id": "delivery_alignment", "label": "Business-delivery alignment", "questionAnswered": "Does the demonstrated reasoning support this delivery-chain node?", "visibleFields": ["outcomes", "decisions", "dependencies", "handoffs", "risk", "validation duties", "escalation", "accountability"], "limitation": "Delivery alignment is configured for a specific company and role, not a universal ranking of people." }, { "id": "method_execution", "label": "Method execution", "questionAnswered": "Which controls actually ran?", "visibleFields": ["method family", "required-input state", "runtime state", "output class", "blocked reason", "version"], "limitation": "A method inventory must not imply that every method applies to every interview." }, { "id": "decision_constraints", "label": "Decision constraints", "questionAnswered": "Can an average hide a critical failure?", "visibleFields": ["must-have gates", "critical-question gates", "anchor-depth limits", "missing-data rules", "recommendation constraints"], "limitation": "A gate is only proven when the released runtime records that it executed under the expected version." }, { "id": "auditability", "label": "Auditability", "questionAnswered": "Can the result be inspected and reproduced?", "visibleFields": ["input versions", "evidence links", "method version", "calculation receipt", "reviewer actions", "corrections", "release state"], "limitation": "Repeatability requires repeated-run evidence under identical locked inputs, parameters, policies, and software version." } ], "evidenceStates": ["supported", "partially_supported", "contradicted", "not_observed", "not_evaluable"], "opportunityStates": ["valid_opportunity", "limited_opportunity", "no_opportunity", "unknown_opportunity"], "ownershipStates": ["personal_ownership_stated", "shared_ownership_stated", "team_outcome_only", "ownership_unclear", "not_applicable"], "runtimeStates": ["computed", "not_applicable", "insufficient_data", "missing_parameters", "unimplemented", "blocked"], "workReasoningDimensions": [ "problem framing", "decomposition", "evidence use", "decision explanation", "tradeoff awareness", "validation and feedback", "adaptation", "ownership boundaries", "delivery coordination", "risk recognition", "communication clarity" ], "terminologyBoundary": { "mentalShape": "Plain-language shorthand for the evidence-bound, role-specific pattern formed by supported work-reasoning observations.", "neuroPsychometricAlignment": "Structured mapping of observable, job-related work-reasoning evidence to a configured role and business-delivery contract, without claiming brain measurement, clinical psychometrics, personality diagnosis, intelligence testing, or access to a private mental state." }, "humanReviewActions": ["confirm", "correct", "exclude", "request_more_evidence", "escalate", "block_release", "release"], "prohibitedInferences": [ "face", "gaze", "appearance", "accent", "voice quality", "emotion", "truthfulness", "mental health", "personality", "intelligence", "disability", "protected traits", "cultural conformity", "hidden mental state", "future performance without a separately established validation study" ], "proofLadder": [ "registry integrity", "execution integrity", "repeatability", "criterion and predictive validity", "fairness and generalization" ], "publicLimitations": [ "The public documentation does not establish quantified accuracy, reliability, validity, fairness, error rate, predictive validity, or improved business outcomes.", "Architecture documentation alone is not a production execution receipt.", "A deployed release must authenticate its method version, provider configuration, and execution receipt before runtime separation is treated as verified." ], "excludedFromPublicResource": [ "formulas", "exact parameters", "weights", "thresholds", "score anchors", "calibration logic", "private prompts", "internal harness instructions", "hidden reasoning", "candidate data", "customer data", "production implementation details" ], "canonicalSources": [ "https://axiomcx.dev/processing-engine/", "https://axiomcx.dev/processing-engine/index.md", "https://axiomcx.dev/knowledge/product-overview.md", "https://axiomcx.dev/knowledge/evaluation-workflow.md", "https://axiomcx.dev/knowledge/evidence-model.md", "https://axiomcx.dev/knowledge/work-reasoning-dimensions.md", "https://axiomcx.dev/knowledge/human-review.md", "https://axiomcx.dev/knowledge/fairness-and-limitations.md", "https://axiomcx.dev/knowledge/public-claim-boundary.md" ], "mathematicalInterpretation": { "title": "From interview answers to role alignment", "summary": "Axiom Cortex connects what the candidate said, how the answer addresses the problem, and what the role requires. The report brings the alignment result, supporting evidence, critical gaps, and follow-up questions into one review.", "sourceBasis": "Explanation reviewed against the Axiom Cortex V4 specification, Calculus Execution Engine design, Scientific R&D report, Field Guide, and local evaluation processing code. Private interviews informed the workflow review; their content is not included in this resource.", "reviewedOn": "2026-09-24", "executionBoundary": "The areas below describe documented methods and their intended interpretation. Each evaluation identifies which configured methods actually computed a result. Research methods require their own data, implementations, and validation; inclusion here does not establish production execution.", "answerFlow": [ { "label": "Lock the role reference", "description": "Bind the job, must-haves, exact questions, and approved ideal-answer criteria before reviewing candidate evidence." }, { "label": "Build each Answer Evaluation Unit", "description": "Connect the question, ideal answer, complete attributable answer, and relevant follow-ups. Preserve the original candidate words and source location." }, { "label": "Separate evidence from context", "description": "Identify demonstrated criteria, partial support, contradictions, and gaps. Interviewer suggestions cannot earn candidate credit. Cross-answer context remains traceable to its original question." }, { "label": "Examine the behavioral axioms", "description": "Review accuracy, mental model, procedural knowledge, clarity, and cognitive load against job-related answer evidence." }, { "label": "Calculate the configured result", "description": "Versioned software applies the selected scoring policy to accepted evidence and approved measurements. Role importance, interview coverage, and critical requirements remain visible." }, { "label": "Review alignment and unresolved gaps", "description": "Read the result alongside the Evidence Locker, must-have coverage, cross-answer consistency, and targeted follow-up questions. The hiring team makes the final decision." } ], "areas": [ { "id": "semantic_alignment", "label": "Meaning and conceptual distance", "methods": [ "Conceptual Fidelity", "Fréchet semantic distance" ], "analysis": "Compare the concepts expressed in the answer with the concepts required by the ideal-answer blueprint.", "interpretation": "Shows where the technical meaning aligns, diverges, or lacks support. Correct paraphrases can preserve meaning without repeating the blueprint.", "executionScope": "Distance methods require approved embeddings, comparison inputs, and a calibrated mapping. Semantic similarity alone does not establish technical correctness." }, { "id": "reasoning_structure", "label": "Reasoning structure and connections", "methods": [ "Discourse analysis", "Optimal Transport", "Wasserstein distance" ], "analysis": "Compare represented relationships among concepts and steps, including the connection between a claim, its mechanism, and its consequence.", "interpretation": "Helps distinguish a connected explanation from adjacent technical terms and locate differences from the expected reasoning.", "executionScope": "A transport distance describes the supplied representations and cost model. It does not prove a reasoning sequence unless those relationships are explicitly represented." }, { "id": "behavioral_axioms", "label": "Five behavioral axioms", "methods": [ "Accuracy", "Mental Model", "Procedural Knowledge", "Clarity", "Cognitive Load" ], "analysis": "Examine correctness, causal understanding, execution steps, explanation clarity, and handling of technical complexity within each answer.", "interpretation": "Shows which part of an answer is supported and which needs more evidence. A correct definition and a demonstrated implementation address different criteria.", "executionScope": "Measurements stay tied to the approved role rubric. Cognitive Load is an evidence-bound rubric dimension, not a measurement of brain activity, stress, or health." }, { "id": "work_reasoning_synthesis", "label": "Depth, adaptation, ownership, and collaboration", "methods": [ "Latent Trait Inference", "Logical Knowledge Depth", "Problem-Solving Trajectory Analysis", "Context Setting", "Metacognitive Calibration" ], "analysis": "Connect answer-level findings to mechanisms, problem decomposition, changing constraints, clarification, contribution, stakeholder impact, and acknowledgment of unknowns.", "interpretation": "Builds the role-specific mental-shape view: architectural instinct, problem-solving agility, learning orientation, and collaborative mindset.", "executionScope": "Trait synthesis needs an approved mapping from answer evidence. The separate six-domain human-task-agent research model must not be substituted for a production scoring policy." }, { "id": "uncertainty_followup", "label": "Uncertainty and the next useful question", "methods": [ "Bayesian evidence update", "Expected Information Gain" ], "analysis": "Model how additional evidence changes an estimate and which approved follow-up could reduce the remaining uncertainty.", "interpretation": "Directs attention to the gap that matters before a decision, including evidence that could change the current interpretation.", "executionScope": "These documented methods require calibrated priors, question mappings, and measurement-error inputs. Missing inputs cannot become an invented confidence interval." }, { "id": "skill_relationships", "label": "Relationships across demonstrated skills", "methods": [ "Gaussian Graphical Models", "Partial correlation" ], "analysis": "Study associations among measured skill dimensions while accounting for other represented dimensions.", "interpretation": "Research can examine which skill measurements connect and where the available evidence is fragmented.", "executionScope": "Requires an adequate dataset and approved model. The V4 specification keeps this in offline validation or shadow research unless those requirements are met; association does not establish causation." }, { "id": "measurement_quality", "label": "Language calibration, reliability, and bias analysis", "methods": [ "Translation invariance", "Inter-rater reliability", "Generalizability analysis", "Expected Calibration Error", "Differential Item Functioning" ], "analysis": "Test whether equivalent meaning receives consistent treatment, whether evaluators agree, and whether measurements vary across questions or relevant study groups.", "interpretation": "Examines the quality of the measurement itself, including unwanted sensitivity to language form and subgroup differences.", "executionScope": "Reliability, calibration, and fairness statistics need suitable study data. Accent, pauses, pronunciation, and protected traits cannot stand in for job-related evidence." }, { "id": "role_alignment_gates", "label": "Role alignment and critical requirements", "methods": [ "Configured aggregation", "Critical-requirement gates", "Human-task-agent alignment distance" ], "analysis": "Combine approved measurements according to the role and preserve the effect of must-have criteria, assessment coverage, and unresolved requirements.", "interpretation": "Shows the supported fit to a particular delivery role and the gaps an overall average could otherwise conceal.", "executionScope": "One complete versioned policy governs a result. The published six-domain distance model studies human-task-agent alignment separately from production aggregation." } ], "readingRules": [ "Read the score together with its question-level evidence, coverage, and critical requirements.", "A question that was not asked creates an assessment gap. It cannot silently become a candidate failure.", "A complete answer that misses a tested criterion differs from missing, truncated, or unattributable evidence.", "A later answer can clarify or conflict with an earlier one; its source and scoring use must remain explicit.", "A baseline-conditioned speech-pattern change is a reason to inspect the recording and ask a targeted follow-up. It does not establish that a candidate used AI or was dishonest.", "Pauses, slower speech, accent, disfluency, code-switching, translation, disability, and L2 or ESL communication require context and cannot stand in for job-related evidence.", "A mathematical result describes its evidence and configuration. The hiring team reviews the unresolved gaps and makes the decision." ] } }