FINAPS is a financial advisory service that provides advice and simulations based on user attributes and statistical information. Its AI engine contains FP expert and psychology knowledge. A short psychological test and questionnaire is done through an interactive user experience with projection-mapped cards and image recognition.
After that, expert knowledge-based AI automatically delivers dialogue scenarios assessing the user′s financial literacy and psychological bias to money to provide tailormade advice to the user
Behavioral economics/psychology AI
and expert knowledge AI
*This process diagram is a simplified flow of the dialogue with FINAPS customers.
FINAPS service explained video
FINAPS service demo video
Knowledge base overview
By adding expert knowledge (PL needs, BS needs),
the AI will learn and expand its knowledge.
Major features
- Question cards to obtain the user′s enneagram profile to form the user′s psychological money image
- Simple diagnosis of money related data based on statistical information similar to the user′s profile
- AI Financial advice based on FP expert knowledge and statistical information fitting the user′s psychological profile
- Innovative user interface using projection mapping and image recognition
- The AI creates the dialogue scenario based on expert dialogue scenario learning
- It is possible to expand the knowledge base by adding expert knowledge
- Household analysis created with qualitative expert knowledge and a calculation logic API
AI System Configuration Diagram
- The Actor Critic Model reads the scenario data in ② and decides the most appropriate response based on the user profile clusters.
- The scenario data is created by learning the Expert Dialogue Corpus (learning data).If there is a similar conversation flow among scenario patterns, it will evaluated if it is better to move to another scenario based on user′s utterance or User Profile / User Clusters information. This makes it possible to go back and forth between scenarios.
- In addition to basic data of sample users the User Clusters information also has attribute values that can be acquired from behavioral psychological questions. This information is compared with the User Profile and the next utterance or action is based on the attribute of the group having the highest similarity degree.