wondering if anyone can help with this…
What is the best way to prepare dataset for Employe Welcome emails. Also what would be the effective way to write prompt and completion for welcome emails and use sender and receiver names and job roles in prompt and completion.
Are there training datasets already available?
Related
I'm a programmer who is about to release an intranet site where apprentices can rate their educators and vice versa. Currently the system is working as planned, however HR wants some way to verify that the users are OK with their ratings. If not, they should be able to unlock their ratings so that the other person has to re-do the rating.
Unfortunately, HR also wants to reduce the amount of logins that these users have to endure. In the worst case scenario, users have to:
Log in to rate the educator/apprentice
Log in to unlock the rating
Log in to rate the educator/apprentice again
And so on...
The user who fills the rating has to be user A, while the user who unlocks or confirms the rating has to be user B. User A can also unlock the rating if they have a correction.
This process has to be done twice - once for the educator, once for the apprentice.
There is usually only one workstation present (factory environment).
Possible solution:
My suggestion is a kind of meeting workflow. One user logs in, clicks a button in the appraisal and the workflow starts. The other user is prompted to log in. This starts a kind of "double session" with both users logged in at the same time. This is a way to verify that both of them are present in a meeting.
This process could be used for multiple ratings at the same time, guiding the users through the process one by one.
HR wants both of them to meet and discuss their ratings.
Are there any security and/or best practice concerns that I should be aware of? The system has to be ready on the first of August, so I'm really hoping to solve this issue as easily as possible. Are there better ways to do this?
It turns out that my idea of logging in two users at the same time wasn't so bad. At first, a regular user logs in and launches the meeting mode. In order to verify that the other user is present, this second user logs in. Both user's data is now stored in the session and the meeting workflow launches, guiding both of them through their appraisals. When the last appraisal is finished, the second user is logged out.
This question is solved.
I have a conversation service with intents and entities. I don't really have a dialog flow as my conversation agent would have to respond to a query which is not tiered or rather have a cadence.
For e.g.
Intent1Intent2...Intent n
These are parallel intents which could respond to a question.
What I get on dialog service is a tiered workflow, Start ->Intent 1 -> Intent2
Without Intent1, I cannot go beyond any of the flows.
Kindly advice if I am missing anything here.
Another query related is does natural language classifier service scale to return a response similar to conversation engine? What I wanted was "what is weather in Los Angeles" and it gives me back information as "Temperature" =>Intent and Entity #City=Los Angeles. Assuming that I have list of cities in my Entity City.
Resolved: RTFM. Dialog indeed has the ability to setup parallel intents. Clicking on + at the bottom kind of setsup parallel intents. CLicking on + sign on the right corner allows you to setup intent cadence. I observe that IBM did a good job of copying of what Microsoft did with Luis here. Looks clean with the User interface. When parallel intents are setup, you text responds back with a % match and with that you can make a genuine decision.
I've developed a python app that registers information from incoming emails and saves this information to the GAE Datastore. Registering the emails works just fine. As part of the registration, emails with the same subject and recipients get a conversation ID. However, sometimes emails enter the system so fast after each other, that emails from the same conversation don't get the same ID. This happens because two emails from the same conversation are being processed at the same time and GAE doesn't see the other entry yet when running a query for this conversation.
I've been thinking of a way to prevent this, and think it would be best if the system processes only one email per user at a time (each sender has his own account). This could be done by having a push task queue that first checks if there is currently an email being processed for this user, and if so, put the new task in a pull queue from which it can be retrieved as soon as the previous task has been finished.
The big disadvantage of this, is that (I think) I can't run the push queue asynchronous, which obviously is a big performance disadvantage. Any ideas on what would be a better way to setup such a process?
Apparently this was a typical race-condition. I've made use of the Transactions functionality to prevent multiple processes writing at the same time. Documentation can be found here: https://cloud.google.com/appengine/docs/python/datastore/transactions
I'm designing a system that will issue commands (via service broker) that a user may wish to undo. My thinking was that I could use a pattern like this:
Create a conversation
Send the "do command" message on the newly created conversation
Keep the conversation open..
If user wishes to undo, send the undo message on the conversation
Close conversation
This keeps messages in order (the undo command cannot appear before the initial command it must "undo"), I recognize that this may keep many conversations open indefinitely as users for the most part will not wish to perform an undo. Would this be a performance problem? Is there an alternative using service broker?
You can hold them open for as long as you like, but you probably don't want them to live forever on the system. Is there a period of time after which you can reasonably infer that you are not going to get an undo? An hour perhaps or by the end of the business day?
You can give conversations a maximum lifetime so you can clear them down once you're satisfied that they are logically committed.
I have Exchange 2010 running and i need to track all the emails. I have made a program to get emails from Exchange and do what i need, but the problem is, that if user deletes email or moves it to archive, i can't access it.
Is there a way to make Exchange to write message tracking log to database? Or at least a way to launch program when email is received?
Is good that you wrote an application to do this, but you shouldn't had. Exchange already has Mailbox Audit Logging features and that what you should use, see Understanding Mailbox Audit Logging. You can configure Journaling which automatically records all incomming and outgoing mails. This is a complex topic and rather than re-invent the wheel, I recommend reading about what is already available.