| chatbot_name |
varchar(1024) |
|
false |
|
|
This name will be used when presenting the chatbot to the students. |
| course_id |
uuid |
|
false |
|
public.courses |
The course this chatbot is appearing on |
| created_at |
timestamp with time zone |
now() |
false |
|
|
Timestamp when the record was created. |
| daily_tokens_per_user |
integer |
|
false |
|
|
The number of tokens a student can use per day. Limits the number of messages a student can send to the chatbot based on the complexity of the messages and responses. |
| default_chatbot |
boolean |
false |
false |
|
|
If true, this is the chatbot that is accessible from the balloon on the corner of the course material. If false, this can be some other chatbot related to the course, such as a chatbot visible within the text of the course material. |
| deleted_at |
timestamp with time zone |
|
true |
|
|
Timestamp when the record was deleted. If null, the record is not deleted. |
| enabled_to_students |
boolean |
false |
false |
|
|
If enabled, students can use the chatbot.. |
| frequency_penalty |
real |
0.0 |
false |
|
|
The frequency penalty parameter for the chatbot. The value should be between 0 and 1. |
| hide_citations |
boolean |
false |
false |
|
|
Whether to show citations to user in the chatbot interface. If true, citations will be hidden on the frontend. |
| id |
uuid |
uuid_generate_v4() |
false |
public.chatbot_conversations |
|
A unique, stable identifier for the record. |
| initial_message |
varchar(32376) |
|
false |
|
|
The message the chatbot will send to the student when they open the chat for the first time. |
| maintain_azure_search_index |
boolean |
false |
false |
|
|
If, true, whenever a page is created, updated, or deleted, the the system will sync its contents to the the Azure Search index. If false, the chatbot will not update the Azure Search index. Note that at the moment the system does not delete unused indexes, so yoy want to optimize index costs, you may want to delete the indexes from the Azure portal. |
| max_completion_tokens |
integer |
600 |
false |
|
|
The max. number of tokens the thinking LLM is allowed to use in generating the response, including output tokens and reasoning tokens. Only used with reasoning models. |
| model_id |
uuid |
|
false |
|
public.chatbot_configurations_models |
The LLM to use in this chatbot configuration. The model choice affects the some of the behaviour of the chatbot. |
| presence_penalty |
real |
0.0 |
false |
|
|
The presence penalty parameter for the chatbot. The value should be between 0 and 1. |
| prompt |
varchar(32376) |
|
false |
|
|
The prompt that the chatbot will use to start the conversation. |
| reasoning_effort |
reasoning_effort_level |
‘minimal’::reasoning_effort_level |
false |
|
|
Controls the amount of effort (time, tokens) used in the reasoning phase of response generations. A lower level means the model is more likely to use less tokens in reasoning (thinking). A low reasoning effort level will likely negatively affect how successful the model is in tasks that require complicated planning, reasoning, or complicated tool use. Only used with reasoning models. |
| response_max_tokens |
integer |
500 |
false |
|
|
The maximum number of tokens the chatbot can output in a response, i.e. maximum response length in tokens. Only used with non-reasoning models. |
| temperature |
real |
0.7 |
false |
|
|
The temperature parameter for the chatbot. The higher the temperature, the more creative the chatbot will be. The value should be between 0 and 1. |
| thinking_model |
boolean |
|
false |
|
public.chatbot_configurations_models |
|
| top_p |
real |
1.0 |
false |
|
|
The top_p parameter for the chatbot. The higher the top_p, the more creative the chatbot will be. The value should be between 0 and 1. |
| updated_at |
timestamp with time zone |
now() |
false |
|
|
Timestamp when the record was last updated. The field is updated automatically by the set_timestamp trigger. |
| use_azure_search |
boolean |
false |
false |
|
|
Whether to use Azure Search for chatbot search. Requires maintain_azure_search_index to be true. |
| use_semantic_reranking |
boolean |
false |
false |
|
|
Whether to use semantic reranking for chatbot search. Requires the feature to be enabled in Azure. |
| verbosity |
verbosity_level |
‘medium’::verbosity_level |
false |
|
|
Verbosity of the generated response, with a higher level meaning that the model is more likely to generate longer responses and vice versa. Only used with reasoning models. |
| weekly_tokens_per_user |
integer |
|
false |
|
|
The number of tokens a student can use per week. Limits the number of messages a student can send to the chatbot based on the complexity of the messages and responses. |