1. Click on the “Launch NeuralSeek” button. You’ll land on the “Configure” tab.
2. Input your Watson Discovery Service URL, API Key, and Project Id. Click “Save” on the bottom of the page. NeuralSeek will test the connection.
3. Click on the “Seek” Tab, and ask a question or two! The seek tab allows you test questions and view responses and confidence levels.
4. Responses can be provided in English, Spanish, French, German, Italain, and Japanese.  Questions do not need to be asked in the same language as the requested response.  EG: You can ask a question in French and request the output in Spanish.
5. Integrate with Watson Assistant. Follow the Directions on the “Integrate” tab to hook NeuralSeek up to your assistant in a few easy clicks.

1. We automatically apply enhancements to the project that you connect to NeuralSeek.  These include search enhancements, as well as modifications to the actual documents stored in Discovery.  For documents obtained by web scrape, NeuralSeek will aggressively cleanse them to remove nuisance text such as banners and cookie warnings – in order to expose more of your company’s relevant information for real-time use.
2. Avoid conflicting information!  if you upload PDF or other corporate docs, be sure to remove old docs that may conflict with the new ones
3. Feedback Helps!  Any time spent curating or training results inside discovery will help expose preferred answers to questions inside NeuralSeek, however we aim to provide a solid baseline with our automated enhancements so that many questions may be answered sufficiently without training.

Most integrations will want to use the “Actions” based extensions.  NeuralSeek may take 7-8 seconds to return sometimes, and may not work for a webhook-based approach, due to the 7 second timeout of Watson Assistant on webhooks.  Watson Discovery response time is the primary driver for overall search time.  Higher tiers of Watson Discovery provide for faster responses.

1. You can still use the custom extension if using Dialogs.  Simply hook an “anything_else” dialog node to push the user query to an action that calls the extension.
2. The “Integrate” tab walks you thru the steps to install the custom extension.
3. You can also use NeuralSeek with dialog-based Watson Assistant.  Set up an action with the NeuralSeek Extension, and then call that from an “anything_else” node in Dialogs.  It is highly recommended that if you take this approach that you manually “null” the returned context variable after you display the response text.  If you do not, in certain situations your assistant will max out the allowed context size and stop responding to user questions until a new session is started.

1.  The quality of the output of NeuralSeek is directly correlated to the quality of the knowlege base documents loaded into Watson Discovery.  If you are getting undesired output from NeuralSeek, investigate the content that you have provided in Discovery.  The more the better.  Ensure the content is not contradictory.
2. NeuralSeek is designed to always give an answer.  This means that NeuralSeek WILL sometimes give a wrong answer. This is an unavoidable consequence of a system designed to always answer.  You can use the Warning & Minimum confidence settings in NeuralSeek to block the system from answering or provide a warning in low confidence instances.
   

1.  NeuralSeek will automatically generate Watson Assistant “Actions”, based on user questions that are asked.   Generally Watson Assistant needs 5 or more user question examples to train on for a high-confidence match to a user query.  When user questions are cataloged by the system, NeuralSeek automatically tries to generate similar worded questions to meet the minium of 5 user examples.  Similar Question generation may take up to 1 minute to show inside the Curate tab after a new user question is logged.
   
2. Select Intents on the curate tab that you want to curate as Actions into Watson Assistant.  NeuralSeek will highlight other intents that you should NOT select because they will share training data and collide.  NeuralSeek attempts to group like questions by keywords, so it is expected that the same questions and answers may be in multiple intents.  You need to choose the grouping that works best for your usecase.  Once you download your Actions file, upload it into Watson Assistant by using the gear icon in the upper right of the Actions screen.  If you already have actions, it is importnat to first download them and upload them as “Base Actions” into NeuralSeek.  NeuralSeek will then preserve all of the existing actions as you export your new actions.