Words are windows into the brain

At Assessment Labs, our vision is to revolutionise integrity risk assessment with scalable, language-based behavioural analysis powered by LLMs.

​

“Instead of subjecting people to endless batteries of tests, we could gain valuable insights from samples of language. Words are windows into the brain. The words that we choose — and how we say them — speak volumes about our personalities and even our mental health. Our thoughts, feelings and behaviours are reflected in language.”
 

- Josh Oltmanns, Washington University Psychologist

​​

​​

It Works Independently of Technology Through Expert-Driven Analysis

​

🔹 Every choice of word reflects thought processes, emotions, and personality traits.

🔹 Founder expertise combines extensive language-based behavioural assessment practice, neuroscience, applied psychology, psychoanalysis, and computer science.

🔹 Proven methodology applied by the founder since 2020.

​

Language is one of the most reliable behavioural signals. By systematically analysing language, it is possible to identify indicators of behavioural risk and other personality traits. This process was manual and expert-driven—technology was not required for the analysis itself. Since 2020, I have been applying this methodology as a multilingual expert assessor at a London-based business psychology firm whose main client is the United Nations. During this period, I conducted behavioural assessments of hundreds of professionals across the globe, using client-defined behavioural indicators and my expertise in psychoanalysis, neuroscience, and applied psychology to deliver accurate, actionable insights. Long before AI, experts could extract meaning from raw speech or text through close analysis of choice of words, linguistic markers, coherence, and emotional valence to infer behavioural tendencies and risks.

​

​"Collaboration among psychologists, computer scientists, and others is essential... effective interdisciplinary collaboration will be necessary to develop LLM-based psychological-assessment methods that will be as effective as science and medicine will need them to be."

 

LLMs for Psychological Assessment: A Comprehensive Overview

ASSOCIATION FOR PSYCHOLOGICAL SCIENCE (September 2025)

​​​​​

​​​​​​

How AI Enhances It

​

🔹 Scalability – Analyses thousands of transcripts in minutes.
🔹 Enhanced Detection – LLMs capture subtle linguistic & psychological patterns.
🔹 Consistency & Objectivity – Standardised, bias-reduced results.
🔹 Multilingual – Direct analysis across languages.
🔹 Seamless Integration – Dashboards, APIs, and automated reports.

 

While language-based behavioural assessment can be performed manually, AI significantly amplifies its scalability, accuracy, and consistency. Advances in Large Language Models (LLMs) have unlocked new possibilities. In January 2023, I began integrating ChatGPT into this methodology. The model follows my guidelines and pre-defined behavioural indicators, while I validate and interpret the outputs. This human-guided hybrid approach ensures scientific rigor, context and ethical applicability. AI enhances, it does not replace expert assessment.

​​

"Psychologists are turning to artificial intelligence to uncover hidden psychological cues in speech, from word choice to tone and pacing. These signals can reveal personality traits and even early signs of mental health conditions, but human clinicians may miss them."

​

​

Your Words May Reveal More Than You Think: AI Shows How.

NEUROSCIENCE NEWS (September 2025)

​​

The Technical Core

 

How It Works 

​​

🔹 Language Capture: AI-agent collects audio data via API integrations or website. 

🔹 AI-Powered NLP: Our custom-trained models analyse subtle linguistic markers.

🔹 Signal Extraction: Markers are mapped against validated behavioural indicators.

🔹 Risk Scoring: Risk scores highlighting behavioural red flags are generated.

🔹 Statistical Safeguards: Statistical anomalies trigger human review.

🔹 Human-as-a-Tool: Experts act as AGI for the AI-agent ensuring ethical oversight.

​​​

​​"Humans should always have oversight and final judgment over any consequential decisions that might be made by an LLM."

​​

​​​​

Why It Works

​​​

🔹 Language is an involuntary behavioural output—it reveals cognitive biases and personality traits that individuals cannot easily mask.

🔹 Research in psychiatry, forensic linguistics, and deception detection has consistently shown that patterns in language correlate with risk behaviours, from fraud to aggression to lack of accountability.

🔹 Unlike self-reported questionnaires, language analysis captures authentic, spontaneous signals, making manipulation difficult.

​​​​

"Language as an assessment tool has several strengths. It is behavioural (objective), natural (valid), and rich, allowing individuals to break free of traditional rating scales. It is also scalable." 

​​

​​​​​

Where It Works

​​​​

🔹 SME Lending: Millions of SMEs remain underserved due to insufficient data. Beyond credit scores, beyond numbers, we turn the invisible into insight.

🔹 Fraud Prevention: Our tools provide an additional defense layer for banks, insurance companies and fintechs, helping flag high-risk profiles during onboarding. 

🔹 Executive Assessment: Corporations spend billions annually on leadership hiring. Our product enables faster, scalable, and more accurate vetting.

🔹 Other High-Risk, Low-Information Contexts: We act like night-vision goggles—revealing risk behaviours as clear signals.​

​​​

"In emergencies or particularly low-resource situations, validated LLMs might provide assessments from language when no other assessment would be available."

​​​

LLMs for Psychological Assessment: A Comprehensive Overview

ASSOCIATION FOR PSYCHOLOGICAL SCIENCE (September 2025)