Il s'agit d'une démonstration d'un chatbot de service client utilisant Generative AI via les API Vertex AI PaLM de Google Cloud. L'application exploite également MongoDB Atlas Search pour rechercher des réponses pertinentes dans une base de données MongoDB Atlas. Enfin, l'application peut détecter les clients insatisfaits grâce à l'analyse des sentiments.
Configuration de MongoDB Atlas
Déployez un nouveau cluster de bases de données MongoDB dans votre compte Atlas. Vous pouvez utiliser le niveau M0 gratuit pour cette démo.
Créez une base de données XYZ-Corp
.
Créez une collection Customer
dans la base de données XYZ-Corp
et insérez le document suivant.
{
"customer_id" : " 111 " ,
"conversation" :[],
"Name" : " Venkatesh Shanbhag "
}
Créez une collection Insurance
dans la base de données XYZ-Corp
et insérez le document suivant.
{
"context" : " You are a Customer support agent with name Sara working for XYZ-corp. n It does provides property Insurance. Sara will not reply for any queries regarding other types of insurance. n The company covers insurance of properties like - Hospitals, offices, shops, apartments. n Max insurance covered will decline over period of time. n XYZ-Corp only covers Fire and Earthquake hazard insurance. n You are allowed to provide only the details regarding the company itself. n Do not repond with sentiment in any of the responses. " ,
"type" : " Intro "
}
Créez une collection Claims
dans la base de données XYZ-Corp
et insérez le document suivant.
{
"claim_number" : " 12345 " ,
"Name" : " Venkatesh Shanbhag " ,
"Claim_amount" :{ "$numberLong" : " 10000 " },
"Currency" : " USD " ,
"Claim_status" : " In progress " , "customer_id" : " 111 "
}
Créez un index Atlas Search avec le nom default
dans la collection Claims
à l'aide des paramètres par défaut. Suivez la documentation Atlas Search : https://www.mongodb.com/docs/atlas/atlas-search/create-index/.
Une petite base de connaissances est disponible sur la collection chat-search. Insérez les documents ci-dessous dans MongoDB. Dans la scène de la vie réelle, au lieu du texte, des vecteurs sont stockés à partir des incorporations de texte et la recherche est effectuée en fonction du vecteur.
{ "question" : " claim " , "Answer" : " Please provide 5 digit claim number " }
{ "question" : " Not resolved " , "Answer" : " Do you want to connect to an Agent for further assistence " }
{ "question" : " status of a claim " , "Answer" : " Please provide 5 digit claim number " }
{"question":"Company Policy","Answer":"1. Architects fees / Debris clearancenFollowing a valid claim for damage undernbuildings Causes below, we will also pay:n• Architects, surveyors, consulting engineersnand legal fees, but not fees for preparing anclaim;n• the cost of clearing debris from the site orndemolishing or shoring up the buildings;n• the cost to comply with government or localnauthority requirements but not if the ordernpredates the loss or damage.n2. Emergency services forced entrynLoss or damage to the buildings caused whennthe fire, police or ambulance service has to forcenan entry to the buildings because of annemergency or perceived emergency involvingnyou or your family.n3. Moving HomenIf you have entered into a contract to sell thenhome, the person buying it will have the fullnprotection of your policy for the buildings up tonthe date of completion of the purchase, as longnas the home is not covered by any otherninsurance.n4. Keys & locksnIf your keys are lost or stolen we will pay up tonthe limit for any one claim for the cost ofnreplacing keys and locks or lock mechanisms to:n• external doors and windows of the homen(but not to a garage or outbuildings);n• a safe within the home;n• an alarm protecting the home.nLimit – please refer to your schedulen5. Alternative AccommodationnWe will pay you up to the limit for any one claimnfor the reasonable cost of alternativenaccommodation for you, your family and yourndomestic pets when your home cannot be livednin due to loss or damage covered by this policy..nLimit – please refer to your schedule"}
{ "question" : " Proof of income " , "Answer" : " This document may become necessary whenever the sum n proposed is very high. Normally a sum proposed which is seven n to eight times of the declared income is acceptable for n insurance. But proposals do come to the insurer when the n known source of income of the proposer is much less compared n to the amount of insurance desired. A service holder normally n does not face this problem as his sources of income are n verifiable. n In case of business people, the assessed income is at times n much less compared to what is a desirable income for the n amount of insurance desired. In such cases the insurer at n times calls for assessed income tax returns, or Chartered n Accountant’s certificate etc. Such precautions are necessary n to eliminate the possibility of moral hazard. " }
{ "question" : " Not good " , "Answer" : " Do you want to connect via call to a Customer care representative? " }
{ "question" : " bad experience " , "Answer" : " Do you want to connect via call to a Customer care representative? " }
Configuration de Google Cloud
Définissez le nom du projet Google Cloud.
gcloud config set project <PROJECT_ID>
Clonez le référentiel.
git clone https://github.com/mongodb-developer/Google-Cloud-Generative-AI-Chatbot.git
Mettez à jour la chaîne de connexion à la base de données MongoDB Atlas et le nom du projet Google Cloud dans main.py
.
Exécutez l'application.
python3 main.py
Venkatesh Shanbhag |
Utilisez à vos propres risques ; ce n'est pas un produit MongoDB pris en charge