Latent Insight2020-01-21T23:23:14+00:00

Project Description

Latent Insight


The employment of web services has replaced personal interaction between us and our clients. These services are generally reliable, cost effective and efficient. Unfortunately, the web interaction comes at a cost: we have lost much of the information available to us through a person-person interaction. Until now, it has been difficult to assess a client’s personality without the employment of lengthy, cumbersome, intrusive and often unreliable additional processes.

image of human head/shoulders made from grey triangle shapes

Latent Insight (LI) applies machine learning to profile a person as they provide information and interact with a web-site while the person is using the site. LI provides a highly-accurate personality profile to bridge the barriers common in web transaction systems without requiring our clients to undertake additional processes.

We receive enriched information about a person who may represent high risk by providing false or misleading information. LI can accurately identify a user’s personality and identify non-human actors (eg. bots). It also assists in identifying those who made previous applications in another name enabling:

  • pre-selection of high-risk applicants for more rigorous scrutiny, and
  • saving time and money while streamlining low-risk clients.

LI works in close to real-time, enabling on-line responses to website users.

Neo4j logo


Increasingly businesses and governments are delivering services through on-line channels. Citizens and clients are accessing websites around the clock and from an almost infinite number of locations. Unfortunately , we are only able to get a very limited view of who we are dealing with.

In a web-based environment we frequently lack information about the user at the other end of our website:

  • Who (or what) is on-line?
  • What are their motives?
  • Is our site delivering services effectively or is it confusing our clients?
  • Does the user represent any risks to our business?


Latent Insight can assist by telling us more about the client through the use of our state-of-the-art Artificial Intelligence engine by assessing how a user interacts with the web-site.


Digital Personality Fingerprinting

  • Establish a digital personality fingerprint based on mouse and/or tablet use
  • Verify identity against a digital fingerprint
  • Detect changes in identity online (in real-time)

Identify Personality Types

  • Identify potential client risk (reliability, potential deceit)
  • Enable real-time responses (activate alternate questions/processes)


  • We use the latest artificial intelligence techniques and tools together from a deep understanding of behavioural psychology to accurately map and evaluate a web-user’s online interactions with your website.
  • Our analysis machine can be deployed as a microservice, requiring minimal changes to your existing platforms.


  • Latent Insight extracts information from the observable web behaviour of the applicant he/she is navigating to and processing the online visa application form.
  • A small section of code is added to your website to enable the tracking.
  • Data is sent to a back-end artificial intelligence system that analyses mouse movements in real time. A unique digital personality ‘fingerprint’ is created.
  • The fingerprint can be used to identify personality types. As it is uniquely associated with a user, it can also be used to assist in managing identity.

Possible Uses

  • Detect Potential Fraud: Detect that a person using your site is not the same person whose digital personality fingerprint was established earlier.
  • Detect Potential Fraud: Identify risks posed by the personality type.
  • Detect User Confusion: Identify components of your website that confuse end-users.

By identifying low-risk clients, streamlined processing is enabled while high-risk individuals can be sent to more intense processing channels. This saves the organisation time and money and makes the interaction with your website a more pleasant experience for the majority of low-risk clients.

Person to person interaction allows us to see some important personality traits of clients. This is usually lost when web-services are used. LI’s ability to evaluate the personality of our clients allows us to better understand our client and the client’s needs.

Identification of non-human users facilitates protection against inappropriate use of web sites. LI can detect when it is dealing with a non-human user.
Where appropriate, the unique personality fingerprint for an individual assists in verifying the identity of a registered system user. Misrepresentation of identity can be readily detected.

LI can assist HR departments by identifying the personality traits of employees and job applicants.

LI supports analysis of websites, identifying sections that are confusing to users.


Latent Insight will:

  • Enhance the security and usability of websites without needing to deploy additional and cumbersome tools and processes.
  • Provide you with valuable personality profile information that can inform your client risk-management processes.
  • Assist in managing user identity.

Business Use Cases

Increasingly many transactions use online only channels. These are more accessible to clients than person-to-person operations.

Unfortunately, the lack of personal contact reduces our ability to assess the client fully as our IT systems do not capture a lot of information a skilled client service officer obtains in face to face transactions.

LI addresses this information shortfall. By analysing how a client interacts with a web-site, LI can provide real-time personality profile information.

Using the personality profile data, client confusion, possible deceit and other characteristics can be identified. This information can then be used to assess the risks associated with a client while they are transacting on-line.

Client interactions generate a unique personality profile. This can be linked to a client’s identity. A returning client should generate the same, unique personality fingerprint. Together with other processes, matching the personality fingerprints will enable more confident on-line interactions with returning clients as we have a powerful but invisible (to the client) identity check capability.

Automated systems that probe websites (eg bots) cannot easily generate a human personality profile in LI. It is, therefore, possible to screen for bots and deny them access.

An organisation that is looking for staff can use LI to identify compatible and incompatible personalities in initial (or subsequent) on-line interactions with potential staff. Applicants for positions that display high-risk personalities can be identified and managed with little additional effort.

Online Demonstration

A demonstration of the solution is available in a online form. You will see the profile result generated by you filling out a sample form.

To participate simply click ‘Request Demonstration’ button and provide your details. A link and one-time-password will be emailed to you.

Request Demonstration


Meet the team behind Latent Insight

Hugo Gamboa
Hugo GamboaBio-Med. Engineer / Data Scientist
Hugo is an Assistant Professor at the Physics Department of FCT-UNL and member of LIBPHYS-UNL research center. He is a founder and president of PLUX, a technology-based innovative company in the field of systems and wireless medical sensors. He is a Senior Scientist at Fraunhofer Portugal where he coordinates the scientific activities of Lisbon office focused on hardware design, signal processing and machine learning applied research and industry technology transfer.

His thesis entitled “Multi-Modal Behaviour Biometrics Based on HCI and Electrophysiology” presents new behavioral biometrics modalities which are an important contribution for the state-of-the-art in the field.

Hugo coordinated as Principal Investigator in several National and European projects.

Based in Portugal

Cátia Cepeda
Cátia CepedaBio-Med. Engineer / Scientist
Cátia is in the last year of her PhD in Biomedical Engineering, entitled “Personality assessment using biosignals and human computer interaction”. She has been working in collaboration with psychologists from the University of Zurich for more than 3 years, believing that our body behaviour can express the insight of our mind’s states, thoughts and intentions.

One of the significant outcomes of her work was the personality prediction using features and behaviours extracted from mouse tracking in the context of an online survey.

She has developed a high degree of expertise in signal processing, HCI, machine learning and data visualization.

Based in Portugal

Marcus Cheetham
Marcus CheethamCognitive Neuroscientist
Marcus is the scientific director of multimorbidity- related research and deputy director of all other research at the Department of Internal Medicine of the University Hospital Zurich. The multimorbidity research focuses on Medical Decision Science and Health Data Analytics, integrating conventional and innovative methodologies, for application in complex dynamic decision making situations.

He is also head of the VR-based Decision and Communication Skills Lab at the Department of Internal Medicine.

His background is as a behavioural neuro-psychologist in the field of neuroimaging, and was previously university Professor of Psychology at the founding of NBU in Seoul, South Korea, and is currently involved in several international projects.

Based in Switzerland