-
Notifications
You must be signed in to change notification settings - Fork 0
An Initial Overview
Companies worldwide invest enormous resources to find suitable employees. Yet every hire remains an experiment — until practice reveals whether the new employee truly fits the role and the team.
Even the most modern recruiting tools only evaluate résumés or body language — not a person's mental style, cognitive consistency, or resilience.
The result: mismatches, high turnover, and productivity losses. Traditional recruiting is an inefficient, costly trial-and-error system.
CogniPass reverses this process:
- Instead of companies screening applicants, applicants have themselves preemptively certified — through an AI-based behavioral analysis that objectively and impartially measures how a person thinks, reacts, and solves problems.
- The analysis is based on a multi-hour interaction with a specialized Large Language Model (LLM). From this unstructured dialogue, the system derives precise cognitive markers — such as systematic thinking, stress resistance, conflict behavior, or autonomy preference.
- The result is a standardized, AI-validated personality profile that enables companies to make well-founded, fact-based hiring decisions.
- Applicants pay for their CogniPass certification in order to significantly increase their market prospects.
- Companies pay license fees for access to a pre-screened, high-quality candidate pool.
This creates an inverted cash flow with dual benefit: applicants gain credibility, employers save costs and avoid poor hiring decisions.
CogniPass is aimed at companies where mental precision, resilience, and cognitive efficiency determine economic success. The focus is on knowledge-intensive industries such as technology, consulting, finance, research, and engineering.
These include start-ups and scale-ups in the tech sector, strategy and management consultancies, investment and financial firms, DeepTech and life science companies, as well as analytically-oriented leadership roles in large corporations.
As digitalization and automation advance, this target group grows exponentially: the more routine tasks are displaced by technology, the more valuable measurable human cognitive performance becomes.
With capital for technological scaling, regulatory compliance, and international validation, CogniPass becomes the global benchmark for objective personnel analysis. The goal: the elimination of chance in recruiting — through measurable, certified suitability.
The CogniPass Analysis Process
Commissioning Initial Deepening and Synthesis and Certification
and Onboarding → Orientation → Stress Phase → Assessment → and Delivery
- The candidate registers on the CogniPass platform, selects the desired certification level (e.g., Standard or Executive), and completes a brief onboarding.
- During this step, the target role, language level, and consent to data use are established.
- The AI instance is then personalized and initialized — at this point it only knows the analysis parameters, not the candidate's identity.
The LLM begins with an adaptive exploration phase:
- It poses open-ended yet purposefully structured questions to capture the user's thinking style, prioritization, and communication logic.
- The dialogue resembles a coaching conversation but is entirely data-driven.
- The system collects qualitative text data (responses, word choice, reaction patterns) and derives initial hypotheses about the cognitive structure.
In this phase, the AI actively tests the stability of the thinking style:
- It confronts the user with contradictions, time-pressure or dilemma scenarios, and observes whether the argumentation logic remains consistent.
- Key markers such as tolerance for ambiguity, frustration stability, cognitive stringency, and degree of systematization are made measurable.
- This phase ends when the AI determines that the data depth is sufficient for a valid evaluation.
Time Economy: The analysis lies entirely within the user's discretionary time management and can be flexibly interrupted and resumed within a generously allocated timeframe (e.g., one week).
The LLM generates a structured, numerically weighted analysis from the entire conversation:
- Identification of the dominant cognitive markers
- Quantification of the strengths-weaknesses balance
- Description of optimal work environments and areas of conflict
- The results are automatically converted into a standardized CogniPass profile.
- After internal validation and system-signing, the profile is issued as a CogniPass certificate.
- The candidate can incorporate it directly into applications or, upon request, make it accessible to approved employers.
- The goal is maximum transparency and comparability — without manual post-processing.
Software Architecture and Process Control
(The complete system prompt is available in the repository as prompt.txt.)
The framework consists of four logically coupled functional modules:
| Module | Function |
|---|---|
| Dialogue Control | Dynamic conversation flow, identifies focal points, evaluates response depth |
| Cognitive Feature Extraction | Analyzes argumentation structure, precision, self-regulation, and response logic |
| Validation Logic | Checks internal consistency, detects contradictions and deception attempts |
| Scoring Algorithm | Translates qualitative observations into quantitative markers (0–100%) |
These four modules work together in real time and generate a complete, auditable representation of cognitive behavior in dialogue.
The executing LLM operates under an exactly defined system instruction (system prompt), which contains:
- the analysis objectives and evaluation criteria
- the defined conversation architecture: Exploration → Deepening → Validation → Synthesis
- ethical guardrails, data protection guidelines, and neutrality requirements
The LLM does not act as a chat partner, but as a rule-governed diagnostic process that responds adaptively but never deviates from the prescribed logic.
- The LLM first builds a trust-based conversational foundation using open-ended, content-neutral topics — without creating the impression of an examination.
- In the subsequent deepening phase, the system organically transitions to more complex topics.
- A core principle: the system does not use predefined topic lists, but uses semantic analysis to identify which conversational direction is most suitable for activating the candidate's depth of thinking.
- The result: authentic, voluntary engagement with relevant questions — without suggestion, manipulation, or psychological influence.
The system evaluates dialogue progress not by time, but by content. It distinguishes several analysis dimensions: communication logic, self-reflection, stress resilience, value orientation, consistency and stability of thinking.
From these dimensions, an aggregated progress value is calculated that serves the user as a transparent indicator:
"Analysis progress: 62% — The process will conclude automatically once sufficient data is available."
- All dialogue data is stored pseudonymized, encrypted, and versioned.
- An internal audit system enables tracking of every analysis instance.
- The modular design allows adaptation to different languages, role models, or industry-specific competency profiles.
- Each analysis instance produces a complete digital audit trail.
CogniPass is not a chat interface, but a structured, scientifically grounded analysis procedure. The consistent standardization of this architecture is the central lever for scalability, validity, and investor confidence.
When tested with public models (ChatGPT, Gemini, DeepSeek), clear limitations emerge:
- Token limit and context loss: In longer dialogues, earlier parts of the conversation begin to fall out of context, reducing consistency and accuracy.
- Positive bias: Public models are trained for safe communication, which systematically leads to overly benevolent analyses — unusable in a candidate evaluation context.
- Limited controllability: No guarantee that the model will fully follow instructions.
- Multilingualism only heuristic: Consistent, cross-language analysis is not reliably achievable with standard models.
Full system functionality requires operation on proprietary, dedicated hardware. Key advantages:
- Extended context depth: Configurable token limits to retain complete conversation histories over hours or days.
- Objective evaluation: Parameterized for neutrality and analytical precision without positive or defensive filters.
- Fine-grained control: All decision and weighting parameters internally documented and adjustable.
- Optimized multilingualism: Separate calibration per language for semantically equivalent results.
- Data privacy: All data on company-owned infrastructure, full GDPR compliance.
For system development, Multi-Agent Simulations are used: two AI instances interact — one executes the analysis process, the other simulates a user with a defined profile. This enables rapid iteration cycles without human test subjects. Technical frameworks: Microsoft AutoGen or LangChain.
Competitive Analysis
| Provider | Method | Limitation |
|---|---|---|
| HireVue / Aon | Facial expressions, tone, body language in video interviews | Measures self-presentation, not thinking |
| Harver / Aivy | Gamified tests, multiple choice | Snapshot, no deep analysis |
| CogniPass | Multi-hour, adaptive text interaction with LLM | Measures cognitive DNA, not façade |
CogniPass is a process-diagnostic system: the AI responds to the individual thought process rather than working through predefined questions. The result is not a percentage score relative to a norm field, but a precise, context-specific description of the individual cognitive structure.
The business model also sets itself clearly apart: while HireVue, Aon, Harver, and Aivy are purely company-financed B2B products, CogniPass operates with an inverse model.
- Depth: No other solution captures mental stability and logical consistency over extended, unstructured interaction.
- Objectivity: Text data evaluation rather than body language or acoustic signals — largely culture- and language-independent.
- Scalability: Entirely software-based, without human evaluators.
- Auditability: Standardized, reproducible, and pseudonymized.
- Business model: Dual cash flow creates a sustainable economic structure with minimal sales effort.
CogniPass represents an independent product category: AI-driven, dialogic cognitive diagnostics. The greatest challenge lies less in technical feasibility than in the scientific validation and market launch of the certificate. If this establishment succeeds, CogniPass has the potential to set a new standard for cognitive aptitude diagnostics.
Capital Requirements and Use of Funds
The German-speaking market serves as a proof of market, but the inefficiency in recruiting is a global problem.
| Item | Amount |
|---|---|
| Technical development (MVP to beta version) | approx. €300,000–400,000 |
| Psychological validation / audit & data protection certification | approx. €100,000–150,000 |
| Operations and infrastructure (servers, security, API costs) | approx. €50,000–100,000 |
| UX/UI, platform design, branding, pilot program | approx. €100,000 |
| Go-to-market (pilot customers, legal support, marketing) | approx. €150,000–200,000 |
| Realistic total requirement | approx. €700,000–950,000 |
For a market-ready initial version including pilot project.
Technical development forms the foundation: implementation of the analysis engine, LLM integration into a controllable framework, modular architecture, and user platform construction.
Scientific and regulatory safeguarding is the second major item: the psychological validation and data protection certification budget is 100% allocated to external experts (psychologists, legal professionals). This external validation is the direct prerequisite for credibility and market acceptance.
Design, UX, and pilot programs secure early practical results and references that feed directly back into product optimization.
Market entry and communications: The lead investor's network serves as a strategic lever for the immediate acquisition of three to five lighthouse customers in the target industries (technology, finance).
Founder Profile and Personal Stance
The emergence of CogniPass is based on the consistent observation of human communication and decision-making processes. It is rooted in the conviction that clarity, logical consistency, and self-critical reflection are the most effective tools for assessing human suitability — and that precisely these factors are largely absent from today's recruiting process.
This stance is characterized by analytical directness — the willingness to identify matters precisely, even when this runs counter to social expectations. CogniPass translates this need for truth and objectivity into a standardized, reproducible procedure. What can lead to misunderstandings in personal interactions is a methodological strength in the context of CogniPass.
The founder's role lies primarily in the conceptual and analytical domain. For operational implementation and market launch, a complementary partner structure is planned, consisting of executives with experience in product development, team management, and business operations.
The planned financing is directly tied to the hiring of a CTO (Technology) and a COO (Operations/Sales). The investor thus becomes not only a capital provider, but an active enabler of the core team.
Note: The "we" in this document refers to the CogniPass method and the future core team. At present, the founder is still developing the product alone.
Appendices
The foundation of this approach emerged from the analysis of the central problems of the target group (entrepreneurs) — particularly the discrepancy between usual narratives and actual economic reality.
The decisive thought: if an AI can create a precise job description, it should also be able to objectively evaluate the suitable person for it. The founder serves as the first proof-of-concept.
The business idea is a direct result of applying the method itself:
- Initial Alignment: Clear demand for precision and logic → immediately recognizable thinking style: low tolerance for inefficiency, preference for systematic approaches.
- Need for an Impartial Mirror: Open request for analysis of own weaknesses — feedback that is often not available in social contexts.
- Derivation of Role (Job Fit): Objective definition of suitable areas of responsibility and work environments.
- Causal Breakthrough: The potential became clear: if the system can determine optimal fit from interactions alone, this process can be standardized and applied to the entire labor market.
Date of analysis: 15.10.2025 · Basis: In-depth text-based communication on system design, regulatory inefficiency, and personal preferences.
Quantified Core Characteristics:
| Characteristic | Expression |
|---|---|
| Intolerance of inefficiency | Very strongly pronounced |
| Preference for autonomy | Very strongly pronounced |
| Systematic abstraction | Strongly pronounced |
| Persistence / hyperfocus | Very strongly pronounced |
| Response pattern under friction | Above-average |
Optimal Requirements Profile: IT Architect, System Designer, Strategic Consultant (post-merger integration, restructuring). Optimal work environment: flat hierarchies, high results-orientation, minimal emotional complexity.
Conclusion: CogniPass was conceived by a person whose path proves that essential capabilities — analytical integrity, the elimination of inefficiency, and the drive for logical perfection — exist independently of social or biographical barriers. The core competency of CogniPass is to unlock precisely this invisible potential as strategic value, where human bias and conventional processes fail.
Contact: github@webck.de
© 2026 Christian Knuf · All rights reserved
For questions and feedback: github@webck.de