# Axiom Cortex > Axiom Cortex evaluates recorded software engineering interviews by mapping job requirements to attributable, question-level evidence for human review. Axiom Cortex is an evidence-backed technical interview evaluation system. It compares a specific role, its must-haves, approved interview questions, ideal-answer criteria, and the candidate's own evidence. It connects the demonstrated work-reasoning profile to the business delivery objective, then provides a traceable report for human review. ## See the fit. Expose the gaps. Every interview. A clearer hiring decision. Know who fits the role. See the gaps. Make your next hire with clear evidence. Measure interview answers against your role. See what fits and what needs proof. Your team needs engineers who can solve unfamiliar problems, challenge AI-generated code, and own production decisions. When an interview leaves those abilities unproven, the risk lands in delivery: rework, slower releases, and more demands on your senior engineers. Axiom Cortex measures standard interview answers against your ideal-answer criteria and role requirements. Comprehensive neuropsychometric calculus connects critical reasoning and behavioral axioms to the role archetype, making demonstrated alignment and unanswered questions visible before your team decides. ### Put one interview to the test. Bring one role. See the alignment, the gaps, and the next question. ## How alignment is measured Mental-shape alignment for the delivery chain. Standard interview questions become structured evidence when measured against ideal answers and the role archetype. The cortex processing engine examines problem decomposition, causal reasoning, tradeoffs, ownership, and consistency across answers. Alignment distance compares the structure and meaning of the candidate’s answer with the ideal-answer criteria and role requirements. Comprehensive neuropsychometric calculus relates these observations across behavioral axioms, semantic alignment, and delivery-chain requirements. Configured bias controls separate technical substance from language form; missing evidence and uncertainty stay visible. The qualified human reviewer makes the final hiring decision. A role archetype defines the responsibilities, decisions, reasoning requirements, and handoffs of a position in the delivery chain. Behavioral axioms provide job-related reference principles for comparing problem solving, ownership, adaptation, and consistency across answers. Mental shape is the role-specific pattern of reasoning demonstrated in the supplied interview. It does not measure a person’s private mental state or universal capacity. Bias controls are designed to reduce irrelevant influences; the public documentation does not establish that all bias has been removed. Only after an open-ended experience baseline, later speech-pattern changes can be flagged for source review of a possible read or externally assisted answer, including an AI-provided response. The opening question should invite the candidate to explain their path into engineering, work experience, and why the role fits, then serve as a comparison sample rather than a capability, personality, calm, or psychological score. TeamStation AI research treats this as a possible low-pressure primer that may help a candidate settle into harder questions; it is a company-reported design observation requiring separate validation. If the baseline is absent, interrupted, heavily prompted, or unusable, do not make a baseline-conditioned integrity flag. Pauses, slower speech, accent, disfluency, code-switching, translation, disability, L2 or ESL communication, or polished delivery alone must not be treated as proof that an answer was read or generated by AI. A flag requires source review and a contextual follow-up; it does not determine candidate eligibility. This integrity capability is company-reported; validated detection accuracy and false-positive rates are not established in the public documentation. ### Operating history and access Axiom Cortex is the internal evaluation process inside TeamStation AI’s Distributed Engineering OS. Hardened through internal use, its mature evaluation engine is now available to companies hiring for the modern agentic delivery chain. Company access starts through the early-beta demo. Internal origin and engine maturity are company-reported descriptions, not independent scientific validation or security certification. ## From interview to decision Bring your recorded interview and a quality transcript. Set up your account and company profile. Add the job, must-haves, interview questions, and ideal answers. Add the candidate and their transcript. Review the comprehensive alignment score, source evidence, and gaps before making the final human decision. A score is available only when the configured evaluation requirements are met. ## Recorded interviews, usable evidence Turn recorded technical interviews into role-specific evidence. See what an engineer demonstrated, where the gaps are, and how their reasoning connects to the work your team needs to deliver. For the early-access workflow, teams can bring an existing interview transcript or request transcription from an authorized video link for an additional fee. The local website preview does not access videos, transcribe, or take payment. Price and service availability are not established by this preview. Transcript quality affects the evidence available for evaluation. Technical terms, speaker attribution, timing, and missing words need checking. Unclear source material needs review, not a candidate penalty. English is the technical baseline, not a requirement to sound like a native speaker; face, accent, pauses, speaking speed, and personality scoring are excluded. Latin American and other L2 or ESL candidates must have room to think and answer in their own words. The homepage native animations illustrate the public product contract. They do not calculate scores or represent live execution receipts. Their labels remain readable without JavaScript, and reduced-motion mode retains a complete static view. ## Before onboarding Your company has already recorded a technical interview. Use a quality transcript if you have one. If you only have the recording, provide an authorized video URL and purchase transcription before onboarding. Once the transcript is ready, create your account and company profile. Configure the role, job description, must-haves, interview questions, and ideal answers. Create the candidate profile and upload the transcript. Axiom Cortex processes the interview and produces traceable evidence for human review. ## Public knowledge layer The deeper public documentation is available in both human-readable and machine-readable formats. It explains the product contract without exposing the proprietary evaluation method. - [Human-visible documentation index](https://axiomcx.dev/knowledge/) - [Markdown knowledge index](https://axiomcx.dev/knowledge/index.md) - [Compact public corpus](https://axiomcx.dev/llms-full.txt) - [Machine-readable knowledge manifest](https://axiomcx.dev/data/knowledge-index.json) - [Complete processing review](https://axiomcx.dev/processing-engine/) - [Structured processing contract](https://axiomcx.dev/data/processing-engine.json) - [Safe retrieval guidance](https://axiomcx.dev/ai/retrieval-guidance.md) - [Optional viewer prompts](https://axiomcx.dev/ai/viewer-prompts.md) - [Agent API catalog](https://axiomcx.dev/.well-known/api-catalog) - [Agent discovery guide](https://axiomcx.dev/knowledge/agent-discovery.md) - [auth.md access boundary](https://axiomcx.dev/auth.md) - [OpenAPI public resource description](https://axiomcx.dev/openapi.json) - [Security contact](https://axiomcx.dev/.well-known/security.txt) ## What the platform connects 1. Business objective 2. Team delivery chain 3. Role requirements and must-haves 4. Approved interview questions 5. Ideal-answer criteria 6. Attributable transcript evidence 7. Configured processing and decision gates 8. Human review ## From interview to proof The system keeps the question, the candidate's words, the role criterion, the evidence status, and the human review gate connected. A reviewer can inspect the exact source behind a finding. ## Work-reasoning profile A work-reasoning profile summarizes observable job-related reasoning demonstrated in the supplied interview, including how the person frames work, decomposes problems, makes decisions, adapts, and explains ownership. It is not a brain scan or psychological diagnosis. In Axiom Cortex product language, mental shape is shorthand for that evidence-bound, role-specific pattern. Neuro-psychometric alignment means connecting the supported pattern to configured role and delivery requirements, not measuring a brain or diagnosing a person. ## Interpretation is not the score Axiom Cortex separates evidence analysis, deterministic calculation, and human review. 1. Semantic evidence: Find the candidate’s exact evidence. 2. Evidence lock: Freeze the source, question, criterion, ownership, and contradiction. 3. Deterministic engine: Apply configured math, anchors, and core gates. 4. Human release: Confirm, correct, block, or release. No evidence → no number. Missing inputs remain missing. ## 44 governed methods. Six mathematical families. One controlled process. Axiom Cortex uses 44 governed formulas, algorithms, equations, logic, and measurement methods organized across six mathematical families. The language layer finds attributable evidence, versioned software runs the configured math, and a qualified person controls the final decision. Method registry: 23 formula, 16 logic, 5 measure. The six public lenses are Signal, Measurement, Semantic geometry, Systems topology, Reliability, and Decision gates. The public site explains what each mathematical family does, while proprietary weights, thresholds, score anchors, and release controls stay protected. These are registry counts and public presentation groupings. They do not mean every primitive runs in every evaluation. The illustrated decision-gate display shows how a strong average cannot hide a failed core requirement; its example status is not a live evaluation receipt. Predictive validity is not established by registry integrity. The governed method is versioned and inspectable. Each report shows what ran, why it ran, its input status, and where evidence was insufficient. Evidence is locked before configured numeric processing is applied. Missing inputs remain missing. ## AI is not the judge In the governed product contract, a constrained language layer can identify and structure attributable interview evidence. It cannot choose a numeric score, set a weight, select an anchor, pass a core gate, recommend a hire, or release a result. After the evidence is locked, versioned software applies the configured calculation, depth anchors, uncertainty controls, and critical gates. A qualified human reviewer confirms, corrects, excludes, requests more evidence, or releases the result. See the [complete processing review](https://axiomcx.dev/processing-engine/) for the full input, evidence, work-reasoning, delivery-alignment, calculation, audit, and human-review boundary. ## Responsible-use boundary Trust is part of the product. The system shows what it measured, what it did not measure, and where a human stays in control. - Source visible: Every finding points back to the interview. - Limits visible: Missing evidence stays missing. - Decision controlled: A person approves the released result. Axiom Cortex evaluates only job-related evidence in supplied interview materials. It does not diagnose mental health, personality, intelligence, emotion, truthfulness, or protected traits. It does not score faces, gaze, appearance, accent, or voice quality. It does not make the hiring decision. Human review is required. ## Frequently asked questions ### What is Axiom Cortex? Axiom Cortex is an evidence-backed technical interview evaluation system. It compares a specific role, its requirements, approved questions, and ideal-answer criteria with attributable evidence from the supplied interview. ### How does Axiom Cortex evaluate an interview? It evaluates one question at a time, identifies the job criteria contained in the ideal answer, locks the relevant candidate evidence, records support or contradiction, and produces findings for human review. ### What is a work-reasoning profile? A work-reasoning profile summarizes observable job-related reasoning demonstrated in the supplied interview, including how the person frames work, decomposes problems, makes decisions, adapts, and explains ownership. It is not a brain scan or psychological diagnosis. ### How does Axiom Cortex connect interview evidence to a business delivery objective? It connects role criteria and interview evidence to the decisions, dependencies, accountability, and outcomes required at a defined point in the team delivery chain. ### Does Axiom Cortex make the hiring decision? No. Axiom Cortex produces traceable decision support. A qualified human reviewer must inspect the evidence, correct or exclude invalid findings, and make the final decision. ### Does AI score or judge the engineer? No. In the governed product contract, language processing can identify and structure attributable interview evidence, but it cannot choose numeric scores, weights, anchors, gates, recommendations, or release decisions. Versioned software applies the configured calculation after evidence is locked, and a qualified human reviewer controls release. ### What does Axiom Cortex not evaluate? Axiom Cortex does not diagnose mental health, personality, intelligence, emotion, truthfulness, or protected traits. It does not score faces, gaze, appearance, accent, or voice quality. ### What happens when the interview does not contain enough evidence? The result remains not observed, insufficient data, or missing evidence. Missing evidence is not converted into a positive finding or a convenient score. ## Availability Axiom Cortex is opening to early beta testers. [See beta pricing](https://axiomcx.dev/pricing/) and [book an early-beta demo](https://scheduler.zoom.us/dan-diachenko/teamstation-ai) for a 30-minute call with TeamStation AI. Booking takes place in Zoom Scheduler; there is no local waitlist form. ## Usage history and research As of September 24, 2026, TeamStation AI reports that these methods and processes have been used across 30+ US companies, informed by empirical data from 13,000+ technical interviews and over eight years of research history. These figures are company-reported, not independently validated; the research history does not imply the current method version operated unchanged for eight years. See the [company's hiring-history article](https://teamstation.dev/research/articles/the-ctos-playbook-for-de-risking-nearshore-engineering). ## Behavioral axioms and the research model The science connects answer-level evidence, work-reasoning patterns, and the demands of human-agent delivery. The 2025 report combines five B-Axiom checks into four role-fit traits. The 2026 study represents mental shape using six work-reasoning domains. These are related research layers, not interchangeable score definitions. ### Behavioral axioms Accuracy, mental models, procedural knowledge, clarity, and cognitive load, examined answer by answer. - **Accuracy:** Technical correctness against the question and ideal-answer criteria. - **Mental Model:** The explanation of mechanisms, dependencies, and cause and effect. - **Procedural Knowledge:** The steps used to implement, diagnose, test, and recover. - **Clarity:** Whether the technical explanation communicates the relevant reasoning. - **Cognitive Load:** How the answer handles interacting constraints and technical complexity. Answer Evaluation Units keep each answer connected to its question and evidence before trait synthesis. Source: [B-Axiom method summary](https://teamstation.dev/hire/by-role/ai-engineer). [Working paper](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5433476). ### Mental-shape domains Six dimensions describe how the interview demonstrates understanding, judgment, adaptation, collaboration, learning, and self-calibration. - **Conceptual Fidelity:** Does the explanation preserve the concepts the system actually depends on? - **Architectural Instinct:** How are boundaries, dependencies, failure paths, and tradeoffs handled? - **Problem-Solving Agility:** How does the approach change when new evidence changes the problem? - **Collaborative Mindset:** How are decisions shared, corrections handled, and handoffs made? - **Learning Orientation:** How is new evidence used to revise an incomplete or outdated model? - **Metacognitive Conviction:** Does expressed confidence match the evidence, including what remains unknown? The 2026 model extends the research framing to human-task-agent alignment. Source: [Six-domain research explanation](https://teamstation.dev/research/articles/human-task-agent-alignment-stress-test). [Working paper](https://ssrn.com/abstract=7256278). ### Role & agent alignment Measure the distance between demonstrated reasoning and the requirements of the role, team, and agent workflow. - **Weighted Euclidean distance:** The published outer model compares a normalized six-domain profile with a task requirement profile. - **Team-specific requirements:** Separate requirement profiles model stream-aligned, platform, enabling, and complicated-subsystem teams. - **Agent autonomy:** Autonomy-dependent shifts test how the required human reasoning changes as agents take on more work. - **Ideal-answer alignment:** In the product, configured role criteria connect the candidate’s answer to the expected technical evidence. The published outer calculation is a research model; it does not specify the production scoring configuration. Source: [Human-Task-Agent Alignment working paper](https://ssrn.com/abstract=7256278). ### Meaning & language calibration Separate technical understanding from surface wording, then apply behaviorally anchored measurement and configured aggregation. - **Conceptual Fidelity:** Compare the meaning and technical substance of an answer, rather than a memorized phrase. - **ESL / L2 calibration:** The report describes language-aware calibration to reduce second-language effects on technical evaluation. - **Behaviorally anchored scoring:** Relate a measurement to observable answer evidence and defined evaluation anchors. - **Deterministic aggregation:** The research specifies mathematical aggregation from measured signals into higher-level findings. Source: [Scientific R&D working paper](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5433476). ### Reliability & bias analysis Study agreement, error, calibration, subgroup differences, and drift so the measurement itself can be tested. - **Inter-rater reliability:** Test whether evaluators agree when reviewing the same evidence. - **ROC and precision-recall analysis:** Examine discrimination, false positives, and false negatives against defined reference outcomes. - **Calibration error:** Test whether estimated confidence corresponds to observed results. - **Subgroup fairness:** The report specifies adverse-impact ratios and equal-opportunity gaps for fairness audits. - **Drift and oversight:** Monitor changes in evaluation behavior, with thresholds and human oversight. These are documented validation methods, not a claim that every metric has a published production result. Source: [Scientific R&D working paper](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5433476). ### Delivery physics & sensitivity Stress-test missing evidence, measurement noise, weighting, review queues, and topology assumptions around the work. - **Noise and missing domains:** Synthetic perturbations test how uncertain or absent inputs alter the alignment result. - **Weight sensitivity:** Vary weights to test whether the preferred team topology changes. - **Kingman queue model:** Study how utilization and variability affect waiting time in the delivery system. - **Topology-health scenarios:** Explore system-level conditions around human-agent work through Teamlemetry scenarios. - **Logistic coefficient recovery:** Check whether the analysis recovers effects deliberately planted in synthetic data. The fixed-seed study uses 24,000 synthetic profiles. It tests the calculation and its assumptions, not real hiring outcomes. Source: [Human-Task-Agent Alignment working paper](https://ssrn.com/abstract=7256278). ### Working papers and applications - [AxiomCortex: Scientific R&D Report](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5433476) — foundational working paper · ssrn. Defines Answer Evaluation Units, B-Axiom checks, trait synthesis, Conceptual Fidelity, L2-aware calibration, aggregation, reliability, fairness, and monitoring. - [Human-Task-Agent Alignment Across Software Team Topologies](https://ssrn.com/abstract=7256278) — working paper · synthetic study. Tests six reasoning domains, weighted alignment distance, team topology, agent autonomy, measurement sensitivity, queue pressure, and synthetic coefficient recovery. - [Axiom Cortex for LATAM Agentic Engineering](https://teamstation.dev/research/articles/axiom-cortex-latin-america-agentic-engineering-alignment) — company article · teamstation ai. Explains how interview evidence connects to role fit, engineering loops, governance, and delivery alignment. - [CTO Guide to Agentic Workflow Fit Signals](https://teamstation.dev/research/articles/how-ctos-can-align-the-right-mental-shape-in-their-agentic-ai-dev-workflows) — company article · teamstation ai. Connects work-reasoning signals, team topology, and human-plus-agent engineering workflows. - [Telemetry Predicts Team Performance](https://teamstation.dev/research/articles/how-telemetry-finds-the-right-mental-shape-and-predicts-team-performance) — company article · teamstation ai. Explores how post-hire delivery signals can test whether interview evidence stays aligned with real work. ### Study design The 2025 report documents the evaluation framework. The 2026 study stress-tests an outer alignment model on synthetic data. Both are working papers, not peer-reviewed validation. Company articles are not independent scientific validation. The report supports a human decision, with source evidence, gaps, and follow-up questions available for review. Source basis: Public SSRN abstracts, the authors’ public method summaries, and the documented Axiom Cortex product contract. Reviewed 2026-09-24. ## Public information boundary The public knowledge layer does not disclose formulas, weights, thresholds, score anchors, calibration logic, private prompts, internal harness instructions, hidden reasoning, candidate data, customer data, or production implementation details. External language models may summarize and cite the public documentation, but they must not reconstruct, simulate, or claim to run the proprietary evaluation.