Feature/diarization v1

apps/AppWithWearable/Tasks.md

@@ -1,20 +1,57 @@
 ### Tasks
 
-- [ ] It shouldn't require to reconnect every time that you open the app, it should load `ConnectDeviceWidget` and in there listen/reconnect to the device.
-- [ ] Device disconnected, display dialog to user asking to reconnect, or take the user back to `find_devices` page.
-- [ ] Settings bottom sheet, improve way of handling `***` blurring of api keys, as if you save it while is blurred with *, it sets the key to that value, and you have to set them again
-- [ ] [iOS] memories and chat page on the bottom do not have the blurred colors pattern, but plain primary color
-- [ ] Improve structured memory results performance by sending n previous memories as part of the structuring but as context, not as part of the structure, so that if there's some reference to a person, and then you use a pronoun, the LLM understands what you are referring to.
+- [ ] It shouldn't require to reconnect every time that you open the app, it should
+  load `ConnectDeviceWidget` and in there listen/reconnect to the device.
+- [ ] Device disconnected, display dialog to user asking to reconnect, or take the user back
+  to `find_devices` page.
+- [ ] Settings bottom sheet, improve way of handling `***` blurring of api keys, as if you save it
+  while is blurred with *, it sets the key to that value, and you have to set them again
+- [ ] [iOS] memories and chat page on the bottom do not have the blurred colors pattern, but plain
+  primary color
+- [ ] Improve structured memory results performance by sending n previous memories as part of the
+  structuring but as context, not as part of the structure, so that if there's some reference to a
+  person, and then you use a pronoun, the LLM understands what you are referring to.
 - [ ] Migrate MemoryRecord from SharedPreferences to sqlite
 - [ ] Implement [similarity search](https://www.pinecone.io/learn/vector-similarity/) locally
-  - [ ] Use from the AppStandalone `_ragContext` function as a baseline for creating the query embedding.
-  - [ ] When a memory is created, compute the vector embedding and store it locally.
-  - [ ] When the user sends a question in the chat, extract from the AppStandalone the `function_calling` that determines if the message requires context, if that's the case, retrieve the top 10 most similar vectors ~~ For an initial version we can read all memories from sqlite or SharedPreferences, and compute the formula between the query and each vector.
-  - [ ] Use that as context, and ask to the LLM. Retrieve the prompt from the AppStandalone.
+    - [ ] Use from the AppStandalone `_ragContext` function as a baseline for creating the query
+      embedding.
+    - [ ] When a memory is created, compute the vector embedding and store it locally.
+    - [ ] When the user sends a question in the chat, extract from the AppStandalone
+      the `function_calling` that determines if the message requires context, if that's the case,
+      retrieve the top 10 most similar vectors ~~ For an initial version we can read all memories
+      from sqlite or SharedPreferences, and compute the formula between the query and each vector.
+    - [ ] Use that as context, and ask to the LLM. Retrieve the prompt from the AppStandalone.
+    - [X] -----
+    - [ ] Another option is to use one of the vector db libraries available for
+      dart https://github.com/FastCodeAI/DVDB or https://pub.dev/packages/chromadb
 - [ ] Settings Deepgram + openAI key are forced to be set
-- [ ] In case an API key fails, either Deepgram WebSocket connection fails, or GPT requests, let the user know the error message, either has no more credits, api key is invalid, etc.
-- [ ] Improve connected device page UI, including transcription text, and when memory creates after 30 seconds, let the user know
-- [ ] Structure the memory asking JSON output `{"title", "summary"}`, in that way we can have better parsed data.
-- [ ] Test/Implement [speaker diarization](https://developers.deepgram.com/docs/diarization) to recognize multiple speakers in transcription, use that for better context when creating the structured memory.
-- [ ] Better `AppWithWerable`  folders structure.
-- [ ] Define flutter code style rules.
+- [ ] In case an API key fails, either Deepgram WebSocket connection fails, or GPT requests, let the
+  user know the error message, either has no more credits, api key is invalid, etc.
+- [ ] Improve connected device page UI, including transcription text, and when memory creates after
+  30 seconds, let the user know
+- [ ] Structure the memory asking JSON output `{"title", "summary"}`, in that way we can have better
+  parsed data.
+- [ ] Test/Implement [speaker diarization](https://developers.deepgram.com/docs/diarization) to
+  recognize multiple speakers in transcription, use that for better context when creating the
+  structured memory.
+- [x] Better `AppWithWerable` folders structure.
+- [ ] Define flutter code style rules.
+- [ ] Include documentation on how to run `AppWithWearable`.
+
+---
+
+- [ ] Multilanguage option, implement settings selector, and use that for the deepgram websocket
+  creation
+- [ ] Option for storing your transcripts somewhere in the cloud, user inputs their own GCP storage
+  bucket + auth key, and the files are uploaded there + a reference is stored in the MemoryRecord
+  object.
+
+- [ ] ~~ (Idea) Detect a keyword or special order e.g. "Hey Friend" (but not so generic) and
+  triggers a prompt execution + response. This would require a few hardware updates (could also be a
+  button on the device), and it's way bigger than it seems.
+- [ ] ~~ (Idea) Store the location at which the memory was created, and have saved places, like "at
+  Home you were chatting about x and y"
+- [ ] ~~ (Idea) Speaker detection, use something like the python
+  library [librosa](https://github.com/librosa/librosa), so that friend recognizes when is you the
+  one speaking and creates memories better considering that. Maybe even later learns to recognize
+  other people.

apps/AppWithWearable/lib/actions/actions.dart

@@ -16,6 +16,7 @@ Future<void> memoryCreationBlock(String rawMemory) async {
   debugPrint('Structured Memory: $structuredMemory');
   if (structuredMemory.contains("N/A")) {
     await saveFailureMemory(rawMemory, structuredMemory);
+    changeAppStateMemoryCreating();
   } else {
     await finalizeMemoryRecord(rawMemory, structuredMemory);
   }

apps/AppWithWearable/lib/app_state.dart

@@ -367,83 +367,6 @@ class FFAppState extends ChangeNotifier {
   set memoryCreationProcessing(bool _value) {
     _memoryCreationProcessing = _value;
   }
-
-  List<String> _testlist = [];
-
-  List<String> get testlist => _testlist;
-
-  set testlist(List<String> _value) {
-    _testlist = _value;
-  }
-
-  void addToTestlist(String _value) {
-    _testlist.add(_value);
-  }
-
-  void removeFromTestlist(String _value) {
-    _testlist.remove(_value);
-  }
-
-  void removeAtIndexFromTestlist(int _index) {
-    _testlist.removeAt(_index);
-  }
-
-  void updateTestlistAtIndex(
-    int _index,
-    String Function(String) updateFn,
-  ) {
-    _testlist[_index] = updateFn(_testlist[_index]);
-  }
-
-  void insertAtIndexInTestlist(int _index, String _value) {
-    _testlist.insert(_index, _value);
-  }
-
-  String _wav = '';
-
-  String get wav => _wav;
-
-  set wav(String _value) {
-    _wav = _value;
-  }
-
-  List<String> _whispers = [];
-
-  List<String> get whispers => _whispers;
-
-  set whispers(List<String> _value) {
-    _whispers = _value;
-  }
-
-  void addToWhispers(String _value) {
-    _whispers.add(_value);
-  }
-
-  void removeFromWhispers(String _value) {
-    _whispers.remove(_value);
-  }
-
-  void removeAtIndexFromWhispers(int _index) {
-    _whispers.removeAt(_index);
-  }
-
-  void updateWhispersAtIndexWithFunction(
-    int _index,
-    String Function(String) updateFn,
-  ) {
-    _whispers[_index] = updateFn(_whispers[_index]);
-  }
-
-  void updateWhispersAtIndex(
-    int _index,
-    String _value,
-  ) {
-    _whispers[_index] = _value;
-  }
-
-  void insertAtIndexInWhispers(int _index, String _value) {
-    _whispers.insert(_index, _value);
-  }
 }
 
 void _safeInit(Function() initializeField) {