Skip to main content
The WeCareRemote AI assistant is designed to remember. Rather than treating each message as an isolated request, the assistant maintains a running understanding of your conversation — both within a single session and across multiple sessions over time. This means you can pick up where you left off, reference something you mentioned days ago, and receive more personalized, relevant responses as the assistant builds familiarity with your needs. There are two layers of memory at work:
  1. Within-session memory — every message in the current chat is kept in the model’s context window, so the assistant understands follow-ups without restating the topic.
  2. Persistent history — when the session ends, the conversation is written to a database. The next time you open the assistant, those messages are loaded back in so the conversation can continue.
The more you interact with the AI assistant, the better it can tailor responses to your situation. Conversation history lets the assistant recall your previously mentioned preferences, circumstances, and past requests — reducing the need to re-explain context every time.

​
Within-session memory

During an active conversation, the assistant keeps the full history of messages exchanged in that session as part of its context. This means:
  • Follow-up questions are understood without re-stating the topic
  • The assistant can refer back to something you mentioned earlier in the chat
  • Multi-step requests stay coherent across several messages
For example, if you ask “What support is available for families?” and then follow up with “What about in my region?”, the assistant understands that “my region” refers back to your earlier context and responds accordingly.

​
Persistent history

Conversation history is stored in a database and persists beyond the current session. When you return to the AI assistant after closing the browser or logging out, your previous exchanges are still available. This persistent storage means:
  • The assistant can reference conversations from previous sessions
  • You do not need to repeat background information you have already shared
  • NGO admins reviewing interactions can access prior exchanges for continuity of care
Conversation history is retained for as long as your organization’s data retention policy allows. Contact your WeCareRemote administrator if you have questions about how long your history is stored.

​
Starting a fresh conversation

If you want to start a conversation without the assistant drawing on previous history, you can begin a new session or conversation thread. Starting a new thread gives the assistant a clean slate — it will not automatically reference past exchanges unless you bring them up yourself. To start a fresh conversation:
1

Open the AI assistant

Navigate to the AI Assistant section from the main menu.
2

Create a new conversation

Use the New conversation button or option in the conversation list. This starts a new thread with no prior context loaded.
3

Begin your message

Type your first message. The assistant treats this as a fresh interaction, unrelated to previous sessions.
Starting a new conversation does not delete your previous conversations. They remain accessible in your conversation history and can be resumed at any time.

​
Privacy

All conversation data is stored on the WeCareRemote platform. Your messages are retained to support the persistent memory features described on this page.
  • Conversation history is associated with your account and is not shared with other users
  • NGO admins may have access to conversation logs for users within their organization, in accordance with their platform permissions
  • If you have questions about data retention policies or want to request deletion of your conversation history, contact your WeCareRemote administrator
Do not share information you would not want stored — for example, government ID numbers or banking details — unless you have confirmed with your NGO that the deployment is configured to handle such data appropriately. For maximum privacy, ask your administrator about the Ollama (local) provider, which keeps conversation data inside your organization’s own infrastructure.

​
How thread IDs work

When you talk to the assistant through the API, each conversation is identified by a thread_id. Passing the same thread_id to subsequent requests continues the same conversation; omitting it starts a fresh thread. 
{
  "message": "What about my situation specifically?",
  "thread_id": "session-abc123"
}
For full details on the request format and example code, see the AI agent API reference.