SensAI is a new solution that BE Networks has developed and is available in Verity 6.4 and higher. Currently, it is an optional capability that is included with the Verity license.

SensAI requires infrastructure for training and running its AI Models. BE Networks leverages TogetherAI for flexibility in selecting models and platforms. TogetherAI is a cloud service focused on infrastructure for training and running Open-Source AI models. No customer network information is sent to Together.AI.

SensAI requires the Together.AI “meta-llama/Llama-3.3-70B-Instruct-Turbo” LLM Model. Details and pricing can be found here: https://together.ai/models/llama-3-3-70b

The API key provided by TogetherAI is needed to allow connectivity to SensAI. Note that the API Key is needed at the time of installation. To view your TogetherAI API Key, login to the site: https://api.together.ai

Below are the steps needed to deploy SensAI:

1. Ensure that you are running the latest version of Satori, presently 2.1.64.112. Satori is the observability VM within the Verity platform. It also includes containers for running SensAI. You will need to download and upgrade Satori to the latest version with a “1” in the second place holder. This version denotes that it has the SensAI containers.
   
   
2. Upload the correct version of Satori to the Software Package Application Packages dashboard in Verity.

   

3. Once uploaded, click the Deploy button. Satori automatically upgrades when it detects the new version is available. This takes about 5 minutes.
   
4. To verify the correct version is running, check the version in the Satori Dashboard under the Verity tab:
   Ensure that the second number in the version is a “1” as shown in the example above. This is the correct version that includes the SensAI containers.
   
   
5. Go to the Administration Dashboard, and select the VNFs tab.
   
6. Under System Applications, click the Add button (+) and select SensAI:
   
   
   
7. Set the IP Address you want the SensAI Container to use along with the subnet mask in CIDR format. Also include the Gateway address so that the container can communicate over the network to your TogetherAI cloud instance
   
8. Click the check to save the settings. The container will be created and startup.
   
   
9. Next, it is necessary to add the TogetherAI API key. To do this, SSH to the Satori VM and re-run the setup script.
   
10. At the prompt, type sudo ./satori_admin.sh
    
    
11. At the Admin menu, select “1” for setup and press Enter:
    
    
    
12. When the setup wizard goes through the process of setting up Satori, press Enter to accept the default settings that are already set when you first installed Satori. When you get to the TogetherAI API key, enter your API Key:
    
13. This will save the settings and restart and start the containers. Once this is completed, run sudo reboot to reboot the Satori VM and reload everything cleanly.
    
    
14. Log back in to Verity, there will be a new icon in the bottom right corner. This is the SensAI icon.
    
15. Clicking the SensAI presents a chatbot dialog box. Use this to interact with SensAI.