Method of generating a secure record of a conversation

1. A method of generating a secure record of a conversation, comprising the steps of:

(a) recording the conversation using a recording device;

(b) creating a section, chunk or other form of fragment of the conversation at the device;

(c) uniquely encrypting that fragment, and also uniquely encrypting subsequent fragments, at the device, so that decryption of any single fragment does not result in the entire record of the conversation being decrypted;

in which each encrypted fragment is sent to a remote server whilst the conversation is continuing and is also locally stored on the device whilst the conversation is continuing.

2. The method of claim 1 comprising the further steps of (a) retrieving from a remote secure server the conversation or a fragment of the conversation; (b) decrypting the conversation or fragment(s) and (c) an end-user determining if the decrypted conversation or fragment(s) are compliant with regulatory requirements or other laws.

3. The method of claim 1 , (i) in which each encrypted fragment is sent to a remote server and is also locally stored on the device until such time as the device receives from the server a notification that a specific encrypted fragment has been successfully received and stored at the server, at which time the device deletes that locally stored encrypted fragment, or (ii) in which each encrypted fragment is sent to a remote server and is also locally stored on the device until such time as the device receives from the server a notification that all of the fragments making up the entire conversation have been successfully received and stored at the server, at which time the device deletes all of those locally stored encrypted fragments.

4. The method of claim 1 in which when the server identifies any fragments as missing or corrupt, then it requests the device to re-send these.

5. The method of claim 1 , (i) in which each encrypted fragment is sent to a remote server continuously or intermittently whilst the conversation is continuing, or (ii) in which each encrypted fragment is sent to a remote server continuously or intermittently after the conversation is concluded.

6. The method of claim 1 in which the device

(a) buffers in memory the output of a microphone that is recording the conversation

(b) compresses using a lossy or lossless compression algorithm each section, chunk or other form of fragment of the conversation;

(c) uniquely encrypts each fragment;

(d) creates a header for the entire record of the conversation, the entire record being made up of a set of uniquely encrypted fragments;

(e) uploads one or more encrypted fragments and/or the header if network connectivity is available.

7. The method of claim 1 , (i) in which each fragment is automatically encrypted with a different encryption key; or (ii) in which a unique initialization vector and symmetric key are generated for each fragment, and the initialization vector and symmetric key are used to encrypt each fragment; or (iii) in which a hash is generated for the successive fragments as a checksum for the original unencrypted successive fragments; or (iv) in which the memory size of each fragment is chosen for efficient transmission over a network and for reducing the risk of fraud.

8. The method of claim 1 , in which the server generates a unique user encryption key for the user/recording device pairing when the user first registers its device.

9. The method of claim 1 , in which a unique initialization vector and symmetric key are generated for each fragment, and the initialization vector and symmetric key are used to encrypt each fragment, and in which the symmetric key is encrypted using the user encryption key and an RSA algorithm.

10. The method of claim 1 , (i) in which the encrypted fragments are stored locally in the recording device creating a queue of encrypted files, or (ii) in which the encrypted fragments are stored alongside metadata containing details of each fragment, or (iii) in which the encrypted fragments are stored alongside metadata containing details of each fragment and in which metadata contains one or more of the following: time stamp, file type, number of fragments, start time, end time, AGC (Automatic Gain Control) settings, user ID, device ID.

11. The method of claim 1 , (i) in which the fragments are transmitted to a remote server whenever a network/internet connection is established, or (ii) in which the fragments are transmitted to a remote server whenever a network/internet connection is established, and in which a background service on the device checks for network/internet availability and the presence of encrypted fragments.

12. The method of claim 1 , in which the encrypted fragments are stored locally in the recording device creating a queue of encrypted files, and in which the queued fragments are transmitted to the remote server even if the recording is not complete.

13. The method of claim 1 , (i) in which the encrypted fragments are stored alongside metadata containing details of each fragment, and in which the server receives the encrypted fragments alongside their metadata, decrypts them using their unique encryption key and stores them into a secure transactional vault, or (ii) in which the encrypted fragments are stored alongside metadata containing details of each fragment, and in which the server aggregates each fragments and their metadata into a unique conversation record and stores the unique conversation record into a unique permanent vault dedicated to the user, or (iii) in which the encrypted fragments are stored alongside metadata containing details of each fragment, and in which the server aggregates each fragments and their metadata into a unique conversation record and stores the unique conversation record into a unique permanent vault dedicated to the user, and in which the server waits for all the fragments to be received in a specific reasonable time before storing them into the permanent vault.

14. The method of claim 1 , (i) in which an alert is raised on the recording device and/or on the server if all the fragments are not received within a specific time, or (ii) in which the server sends an acknowledgement receipt to the recording device that all the fragments are received, and the fragments are deleted on the recording device when the recording device receives the acknowledgement receipt, or (iii) in which the server sends an acknowledgment receipt to the recording device once a fragment is received, and the received fragment is deleted on the recording device when the recording device receives the acknowledgement receipt.

15. The method of claim 1 , (i) in which the user of the recording device is able to track or get notification on the status of the conversation recording, or (ii) in which each encrypted fragment is sent to a remote server whilst the conversation is continuing and is also locally stored on the device whilst the conversation is continuing, and in which the server is able to track or get notification on the status of the conversation recording.

16. The method of claim 1 , (i) in which the user is allowed to pause uploading activities for a configured maximum duration to conserve network bandwidth, or (ii) in which the recording device only sends the fragments to a remote server if connected to LAN or WiFi to avoid data charges over a cellular network, or (iii) in which the recording device only sends the fragments to a remote server at specific date and times.

17. The method of claim 1 , (i) in which the conversation is immediately retrievable and accessible by an authorised end-user once it has been uploaded on a remote server, or (ii) in which AGC (Automatic Gain Control) data is captured at recording time in order to detect the status of a recording settings, so that potential AGC problems can be reproduced or emulated and debugged, or (iii) in which a server-side speech and language analytics engine processes the conversation recording, or (iv) in which a client or device-side speech and language analytics engine processes the conversation recording to, for example, detect silence, noise, reverberation, poor microphone positioning and uploads this as metadata to a remote server.

18. A system for generating a tamper-proof record of a conversation, the system comprising:

(a) a recording device including a first non-transitory storage medium, the recording device configured to record the conversation on the first non-transitory storage medium; the recording device further configured to, during or after the recording, continuously or intermittently send successive sections, chunks or other form of fragments of the audio recording to a secure remote server,

(b) a secure remote server configured to receive the successive fragments of the audio; and

(c) a secure vault including a second non-transitory storage medium, the secure vault configured to store the successive fragments of the audio on the second non-transitory storage medium; and

in which the recording device is configured such that each encrypted fragment is sent to the remote server whilst the conversation is continuing and is also locally stored on the first non-transitory storage medium whilst the conversation is continuing.

19. A recording device configured to generate a secure record of a conversation, wherein the recording device includes a non-transitory storage medium, the recording device being configured to:

(a) record the conversation;

(b) create a section, chunk or other form of fragment of the conversation at the device;

(c) uniquely encrypt that fragment, and also uniquely encrypt subsequent fragments, at the device, so that decryption of any single fragment does not result in the entire record of the conversation being decrypted; and

in which the recording device is configured such that each encrypted fragment is sent to a remote server whilst the conversation is continuing and is also locally stored on the non-transitory storage medium whilst the conversation is continuing.