Knowledge and Memory
How the AI remembers your conversations and searches through your documents.
The big picture
There are two ways the AI in Opisense stays smart: memory and knowledge. They sound similar, but they work differently — and understanding the distinction helps you get the most out of both.
Memory is conversation context. It's what the AI remembers from your past chats — your preferences, decisions you've made, things you've told it. Memory makes the AI feel like it actually knows you.
Knowledge is uploaded information. It's the documents, files, and data you add to your knowledge base. Knowledge makes the AI an expert on your team's specific content.
How the knowledge base works
The knowledge base is where you store documents that the AI can search through. The process is straightforward:
- Upload — Add documents in common formats (PDFs, Word docs, text files, and more)
- Processing — Opisense reads and indexes your documents automatically so the AI can understand them
- Search — Ask questions in natural language and get answers pulled directly from your documents
You don't need to organize files into folders or tag them. Just upload, and the AI handles the rest. When you ask a question, it searches across everything and pulls the most relevant information.
For complex questions, Opisense uses deep search — a more thorough process that examines multiple documents and cross-references information before giving you an answer. It takes a bit longer but delivers more comprehensive results.
How memory works
Every time you chat with your assistant, it remembers the conversation. This means:
- You don't have to re-explain context from previous sessions
- The AI picks up where you left off
- Your preferences and working patterns are remembered over time
- Important decisions and instructions carry forward
Memory is personal. Your assistant remembers your conversations, not your teammates'. Each person has their own private conversation history.
Agent memory
Agents have their own memory too, but it works a bit differently. An agent's memory is shaped by everyone who interacts with it. If three people chat with your Support Agent throughout the day, the agent remembers all three conversations.
This is by design — agents serve the team, so their memory reflects the team's interactions.
Key terms
- Memory — The AI's ability to recall past conversations and preferences. Persists across sessions so you don't have to repeat yourself.
- Knowledge base — A searchable library of documents you upload to Opisense. The AI uses these to answer questions about your team's specific content.
- Document — Any file you upload to the knowledge base. Opisense processes it automatically so the AI can search through it.
- Context — The combination of memory, knowledge, and current conversation that shapes the AI's responses. More context means more relevant answers.
In practice
Here's how memory and knowledge play out in daily use:
Memory in action: You tell your assistant "I prefer bullet points over long paragraphs." From now on, it formats responses with bullet points. Next week, you ask it to summarize a project update — it uses bullet points without you asking again.
Knowledge in action: Your team uploads the employee handbook to the knowledge base. A new hire asks the assistant "What's the policy on remote work?" — and gets an accurate answer pulled directly from the handbook.
Both together: You've been chatting with your assistant about a client project for weeks (memory). You also uploaded the client's contract to the knowledge base (knowledge). When you ask "Can we extend the deadline?", the AI checks the contract terms and considers what you've discussed — giving you a grounded, context-aware answer.