Skip to main content

Search Your Memory

Find anything you've saved using AI-powered semantic search

Hayden avatar
Written by Hayden
Updated over a week ago

## The Power of Intelligent Search


Unlike traditional search that only finds exact words, Intangle uses **AI-powered semantic search** to understand **meaning**.


This means:

- Search "authentication issues" → finds "login problems"

- Search "meeting notes" → finds "discussion summaries"

- Search by **concept**, not just keywords


## How to Search in the Intangle Web App


### Ask the Assistant


Simply ask what you're looking for:


**Example searches:**

- "What did I save about authentication?"

- "Show me my tasks for the project"

- "Find my notes from last week's meeting"

- "What do I know about React hooks?"

- "Search for anything about the API redesign"


The assistant will search and show you relevant results. Click any result to see full content.


## How to Search from External Tools


The real power is searching from **any connected tool**:


### From Claude Desktop


```

"What's in my Intangle memory about TypeScript?"

"Search my work space for client requirements"

"Show me recent tasks I added"

```


### From ChatGPT


Remember to select Intangle at the start of the chat, then:


```

"Find my Intangle notes about the API"

"What tasks do I have in my work space?"

"Search for anything about the database migration"

```


### From Cursor (while coding)


```

"What did I save in Intangle about authentication?"

"Check my Intangle memory for the API endpoint"

"What tasks are pending in Intangle?"

```


## Understanding Search Results


When you search, Intangle returns all three memory types:


### Context

- General information, facts, notes

- Ranked by relevance to your query

- Shows title, content preview, and topics


### Tasks

- Actionable items you've saved

- Shows status (pending, in progress, completed)

- Shows priority (urgent, high, medium, low)


### Processes

- Structured workflows and procedures

- Shows the steps in order

- Helps you remember "how to do X"


### Search Confidence


The system shows how confident it is about the results:

- **High confidence:** Very relevant matches found

- **Medium confidence:** Some relevant matches

- **Low confidence:** Might need to refine your search


## Search Tips for Best Results


### Be Descriptive


Instead of: "API"

Try: "API authentication endpoint for user login"


### Use Natural Language


Instead of: "react typescript"

Try: "What did I learn about using TypeScript with React?"


### Search by Concept


- "Security vulnerabilities I need to fix"

- "Client feedback about the design"

- "Performance optimization ideas"


### Search by Time (in natural language)


- "What did I save yesterday?"

- "Show me recent tasks"

- "Old notes about the project"


### Filter by Space


- "Search my work space for meeting notes"

- "What's in my personal space about fitness?"


## What Makes Search Powerful?


### Traditional Search (keyword matching)


Search "login issues" → only finds items with those **exact words**


### Intangle Semantic Search


Search "login issues" → finds:

- "Authentication failing for mobile users"

- "Users can't sign in on Safari"

- "Password reset flow is broken"

- "Security token expiration problems"


The AI understands that all these **relate** to login issues, even though they use different words.


## Real-World Search Scenarios


### Scenario 1: Quick Reference


You're in a meeting and someone asks about the API rate limit.


```

In your phone's ChatGPT: "What's our API rate limit in Intangle?"

```


Result: Instant recall without digging through old notes


### Scenario 2: Context Switching


You're switching from one project to another.


```

In Claude Desktop: "Show me everything about Project Phoenix"

```


Result: All relevant memories surface immediately


### Scenario 3: Task Management


You need to see what's on your plate.


```

In Intangle app: "What tasks are pending?"

```


Result: Prioritized list of all open tasks


### Scenario 4: Research Synthesis


You've saved many articles about a topic.


```

"Summarize what I know about GraphQL"

```


Result: The AI synthesizes all your saved information into a summary


## Verify Search Works Everywhere


### Prove it to yourself:


**Test 1: Search from different platforms**

- Save something in the web app

- Search for it from Claude Desktop

- Search for it from ChatGPT

- All should find the same information


**Test 2: Semantic understanding**

- Save: "The database password is reset monthly"

- Search: "when does the password expire?"

- Intangle should find the relevant memory


**Test 3: Multi-item synthesis**

- Save several memories about the same topic

- Ask: "What do I know about [topic]?"

- The assistant synthesizes all related memories


## Why This Changes Everything


### Before Intangle


"What was that thing we discussed about authentication?"

- Search through chat history

- Search through notes apps

- Search through emails

- Give up and ask the team again


### With Intangle


"What did we discuss about authentication?"

- Instant retrieval from any connected tool

- All relevant information in one place

- No context switching required


Your memory becomes an extension of your mind, accessible from anywhere.


## Advanced Search Features


### Topic Filtering


If you've tagged memories with topics:


```

"Search my work space for memories about 'authentication' topic"

```


### Fetching Full Content


Search results might show summaries. To get full content:


```

"Show me the full content of that memory"

"Fetch the complete details for that task"

```


## Tips for Success


- **Start simple:** Try simple searches to get comfortable

- **Same search, different tools:** Experience true cross-platform search

- **Natural language:** The AI understands context

- **Specific vs broad:** Specific queries for specific info, broad queries for exploration

- **Rephrase if needed:** If results aren't relevant, add more context

- **Remember:** Semantic search finds related concepts, not just exact words


---


**Next:** View Auto-Generated Topics | Back:** [Save from External Tools


Did this answer your question?