Category: Chat

Why use Slack for internal IT Support

According to Enlyft, there are around 135,711 companies that currently use slack as collaborative software.

Chances are, your business might be one of those 135,711 companies. Though using slack as a means of communication between members of the same team or different teams in a company leads to productive work culture, companies can leverage the instant messaging platform even more. By turning slack into your business’ employee support hub.

Here’s why.

1. Helps you get more out of your existing software

Be it Jira Cloud for getting notifications about issues as and when they get mentioned, or HubSpot for your sales and marketing efforts, or Freshservice for managing your IT services, you can add the slack version of the apps your business already uses in the channel of your choice from the slack app directory.

This makes it easier for your employees to use slack’s version of IT service management apps like Freshservice for

• 1. Converting slack direct message (DM) conversations to tickets in Freshservice
• 2. Adding slack conversation as ticket description for context
• 3. Getting in touch with the Subject Matter Expert (SME) of an IT issue right away via slack DMs.

And that’s just the tip of the iceberg.

2. It’s super-flexible

With slack, you can create separate channels for different teams or, you can create a shared channel for your organization and a third-party vendor to collaborate. Channels are a great way of letting employees stay in the loop of things. Be it support agents getting notified in slack about tickets raised via DM or a channel, or an agent reacting to an issue raised by a requestor in a channel to let them know about its status, slack helps in streamlining IT processes.

Moreover, your company’s help desk agents will be able to guide an employee with their issue via slack’s audio and video calls feature on-the-go. Pretty Neat, huh?

3. Prevents Silos among Teams

According to this infographic by the Queens University of Charlotte, 49% of Millennials (who are soon to comprise the majority of the workforce) support social tools for workplace collaboration.

In fact, about 40% of Millennials would even pay out of pocket for social collaboration tools to improve productivity. That’s how much they believe in the dream team of collaboration and technology.

Slack, through its plethora of real-time collaboration and document authoring tools like Google Drive, Dropbox or Microsoft OneDrive makes the dream team of collaboration and technology all the more, real. Through pinned messages and files, new joiners or people already in your team’s Slack channel will be able to review things such as:

• Product spec
• Tech spec
• Designs

This way, lots of onboarding meetings and a slew of forwarded email threads to root through for new employees can be avoided.

Team members will also be able to scan through all previous conversations, decisions and the people involved in the channel. This kind of transparency ultimately leads to the reduction of help desk tickets raised by employees. Now that’s more like it.

4. Easier for triaging issues

Let’s say you have a slack channel called #support for employees to report their IT issues. Whenever an employee reports their issue in the #support channel, the employee or the help desk manager can call the attention of another member who might be an SME for that issue by including an @mention. The person mentioned will then be notified immediately. As seen in the above example, the @mention comes in handy when the help desk manager wants to triage issues mentioned in the channel to the corresponding SME. The SME can then update the status of the issue in the channel itself with the help of emojis that’s easier to track for the employee who reported the issue as well as for the help desk manager.

But with limitations.

There are lot of benefits of using slack in your company. And although it is beneficial for companies to use slack, it comes with several limitations especially if you want to automate user interactions and workflows for IT helpdesk or workplace support. Thankfully, you can overcome these limitations by having a slack chatbot that can do much more than just notifications or converting DM to tickets.

What is Slack Chatbot

Slack is a modern workplace communication or chat tool that helps users connect and collaborate quickly. It also serves as a platform for users to interact with 3rd party apps, which makes it easy for all teams to get work done quickly.

With slack being one of the most widely used communication hub, having a chatbot on slack makes it easy for companies to automate user interactions and workflows that help offer several benefits. Companies are starting to adopt bots on slack that can help automate IT Helpdesk and HR support but also see increased adoption in other parts of the business such as finance, and facilities.

Today, several conversational ai no-code platforms offer quick and efficient way for companies to build bots (for different use cases) and deploy them on slack thus making a slack chatbot more appealing as it does not require human intervention and can autonomously support users. And it doesn’t stop there, you can also automate app workflows from your slack chatbot using no-code platforms. And what this means is, slack Chabot can automate both conversations and interact with apps to automate workflows as well that makes it very attractive to deliver customer and employee service automation.

How to build slack bot using a no-code platform.

https://youtu.be/YRhHicf-CQM

Think of workativ assistant like an icing on the cake to all that you can already do with slack at workplace. With workativ assistant, you can build a conversational bot that sounds humane and solves your employees’ issues with the power of both NLP and workflow automation in the snap of a finger.

Yeah, you heard it right. But Thanos’ snap can only turn half of the universe’s population to dust, while a bot can turn most of not all of your employees’ IT woes into dust in no time. Let’s see what makes workativ assistant’s bot so worthy of wielding the infinity gauntlet more than Thanos.

Here is a short video of how to use a no-code platform like workativ assistant to build your slack chatbot.

Get — Free IT Slack Chatbot

How slack chatbot can automate your IT Helpdesk

Slack chatbot can be very useful for companies to automate IT Helpdesk operations. Having an effective slack bot as your first-line agent can really benefit companies to automate repetitive issues, decreasing mean time to respond, improving employee experience and reducing costs.

Let’s look at the top 6 issues that slack bot can automate for IT Helpdesk.

1. Unlocking Accounts

Problem:

Your account has been locked unexpectedly, and you want to get it unlocked quickly! This is one of the most frequent queries that hit the help desk. It is repetitive and has a defined process. A perfect candidate for your slack bot.

Solution:

In this case, the user types in something like Unlock My Account or select Unlock Account from a list of options the Slack chatbot provides. The bot understands the query, asks for user details to verify, and unlocks the account in seconds.

2. Resetting Passwords

Problem:

Password Reset is the most common request in the IT helpdesk. An agent may take longer than a chatbot to resolve such a trivial issue. With the self-service slack chatbot, you can go through the transparent resolution process yourself and get your password reset in seconds.

Solution:

The user has to simply type in Reset Password into their Slack window. The Bot understands and asks the user for some details like email address for verification. Then, it either sends a password reset link to the user’s email address or asks the user to key-in the new password and resets it.

3. User Provisioning

Problem:

When a new user joins a company or moves to a different department, your IT and HR teams works with the process and systems to create new year, assigns email, adds role and user access privileges to various applications and so on depending on company policy. This takes time.

Solution:

Let the Slack chatbot do that for you in seconds!

All you need to do is type in their query or request (say, create an account for Aaron Smith) in the Slack chat box. The bot looks for key words in the user’s input and responds with either more options to clarify the issue or asks for specific user details for verification. Once verified, it proceeds with the provisioning or deprovisioning process as requested.

4. Access Requests

Problem:

Another candidate for automation is access requests when users change roles or needs access to app. The process and efforts are time consuming with traditional IT Helpdesk.

Solution:

Slack bot can handle this easily. A user types in the query (say I am not able to access the HR portal) and the chatbot asks for the email address for verification and verifies if the user has access before. Post validation, the bot either raises a request to the management for providing access to the user or allows access to the particular resource (here HR portal) itself. Takes only a few seconds.

5. Asset Request

Problem:

Companies provide workplace assets for employees. And when the asset breaks or needs new or additional assets like laptop or keypad, employees raise new asset requests with IT Helpdesk. This is a defined process that can be automated fairly quickly.

Solution:

With slack bot, user types in the query (say My laptop is not working) and the bot asks for more detailed description of the issue. When the user elaborates the issue, the bot proceeds with a request to the management for issuing a new laptop, creates an IT ticket and shares it details with the user as well.

6. Suspend a User

Problem:

Suspending users refers to revoking a user’s access to a particular service and removing them from the company’s email list. Businesses come across similar situations quite often. You wonder if that can be done in seconds?

Solution:

With slack bot, user needs to do is type “Suspend < user’s name >” and the bot, after validating and getting required approval, will either suspend the user itself or raise a raise a ticket for second level agent to suspend the user. `

Top 5 reasons for you to use a no-code platform to build your slack chatbot.

Because it’s easy. Here are some reasons on why you should choose a no-code platform to build you slack bot. We have used workativ assistant as an example.

1. One-click slack integration

Deploying a bot in your preferred slack workspace’s channel is as easy as pie. Just click on the add to slack button, grant the permissions required by workativ assistant to access your slack workspace, add the bot to the channel of your choice, and you’re good to go.

2. Automates repetitive IT issues with workflows and not just conversations

According to INRY, password resets constitute an astonishing 20% to 50% of help desk calls. And like you’d already know; password resets aren’t that much of an issue to call another person for a solution when we ourselves can solve it.

By automating repetitive, mundane tasks like a password reset, unlocking an account, etc., the average cost-per-call to your internal help desk gets reduced drastically. And workativ assistant makes it possible using its workflow automation feature to integrate your bot with several IAM apps and create workflow or an automation in just a few clicks of a button for resetting a password. No coding knowledge required. Handy, right?

3. Manage several simple or complex use-cases with chatbot builder

Yes, you can create automations like a breeze using workativ assistant. But what use are those automations when you don’t deliver the solution properly?

This is where workativ assistant’s intuitive chatbot builder comes into play. With chatbot builder’s myriad of options, you can create conversations for any number of scenarios with a human touch more so that people who chat with the bot can’t tell the difference between chatting with a human and a bot. Again, no coding knowledge required.

4. App integrations and workflow marketplace

You can connect your bot with your business’ apps and use pre-built app workflows for your chatbot

Be it sending an OTP via Twilio, creating a ticket via Freshservice, or sending a message to a slack channel, workativ assistant can do this and more for you once you connect it to the required app.

And you don’t even have to create Automations from scratch. Through the workativ assistant workflow marketplace, you can add ready-made automations for the required apps with just the click of a button.

5. Trains itself to better understand your employees

Workativ assistant’s training module enables the bot that you create to learn from its historical data of conversations with your employees and train itself to better detect an employees’ intention for initiating a conversation.

Top 6 benefits of slack bot at workplace using workativ assistant.

Now that you know why using slack as an internal help desk can help transform your workplace support, why not combine it with workativ assistant to let your employees get the best out of your internal IT support?

Think about it. With slack’s superior features and workativ assistant’s conversational ai-powered automations, you can do so much more for your employees.

1. Supercharge self-service for your employees

By deploying a bot built using workativ assistant in the slack channel for reporting IT issues, your employees can solve their problems themselves from the comfort of the channel.

2. 24×7 availability

Help desk agents can’t be available 24×7 to help out employees with their issues. After all, they’re human too. They need ample rest just like the rest of us.

To “help” your help desk agents, you can build a bot that offers IT support for repetitive tasks like a password reset, unlocking an account, etc., (as discussed earlier) in your workplace support slack channel as well as train itself to resolve problems based on solutions previously offered for similar problems by a help desk agent. This allows employees to solve their problems quickly as and when they arise without the hassles of long waiting queues.

3. Call deflection

Enabling employees to resolve their IT issues themselves via bot in your workplace support slack channel prevents the bottleneck of a single help desk agent handling huge number of calls. Providing self-serve to employees via bot also thwarts the need for help desk agents to provide in-person support.

4. All-encompassing

Adding the slack version of your business’ apps from the slack app directory every time can be a bit frustrating.

With workativ assistant, you can connect your business’ apps with bot and automate tasks for those apps instead of adding them from the Slack App Directory. This makes it easier on your IT Admins as well as your employees. Be it creating a ticket in Zendesk or sending an email via Outlook, your employees can use the bot to take care of it.

5. Productive and satisfied employees

Enabling your employees to solve their IT issues themselves in a slack channel via bot leads them to solve their issues then and there itself and continue with their work.

For employees, this eliminates the frustration that comes with the labyrinthine way of getting an IT issue resolved like sending multiple emails, redirects to different SMEs, etc., Employee self-service also let’s help desk agents focus on resolving the more pressing IT issues at hand.

6. Cost savings

With bot deployed in a slack channel, you can get rid of the overheads that come with repetitive tickets and truly see your help desk costs take a dive immediately. Building and using a slack bot with workativ starts at $299 per month flat. Compare that with hiring a live agent. The cost savings are 10x. More on pricing here.

Steps to deploy your bot on slack for Internal IT Support

Deploying your bot on slack is easy. It is just an easy set of steps to follow and you will be done with having a bot on slack in no time. Let us see some of the major steps involved in the process.

Once you have built your bot, and ready to deploy on slack, you can login to your workativ assistant workspace.

• Select Slack as the desired Chat Channel
• Click Add to Slack to start a new integration
• Sign in to your Slack workspace on the URL: https://api.slack.com
• Create a new app for the Workativ Assistant chatbot and select the desired App Name (Bot name)
• Customize your bot by selecting its Logo, Background Colour, and Bot Name
• Update the required information in both Slack App page and Workativ’s Slack Integration page
• Save the changes and click Add to Slack

There you go! You have successfully deployed your chatbot on slack, and you can easily access it by signing into your slack workspace.

Conclusion

Since you have reached the conclusion, you must be clear by now that Slack is one of the best enterprise-level communication tools for the interaction between the chatbot and the users. It is found that 70% of users prefer using slack for reporting their support issues. And while it’s great for people to interact, using Slack chatbot can really transform workplace support by automating both interactions and as well workflows.

In a nutshell, it is evident that companies are ought to select slack as the suitable platform for deploying their chatbots. With the AI-powered no-code platforms like Workativ assistant, you can leap your slack bot journey. Bringing together Workativ Assistant, Slack, and other third-party tools makes your BOT no less than a SUPER-BOT. Power-packed and Smart!

Can’t wait to transform your business’ slack channel into an IT help desk? Get in touch with us, and we’d gladly give you a demo.

Disclaimer: This article was originally published here.

---

Step-by-step Guide to Building Slack Chatbot for Internal IT Support [No code] was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

The Chatbot Conference is only 3 weeks away, and right now, we are having our last sale of the year!

You can save up to $250 on tickets.

And it gets better. We just launched our Points Program and a Community, so every dollar you spend on tickets earns you a point you can use towards future Conferences, our Certified Workshops, and Membership Packages.

Register Today and see you on November 15th.

Cheers!

Stefan

---

Flash Sale! Save up to $250 on Chatbot Conference Tickets. was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

Activation does it means activating your car with a click ( if it has that ,of course) , well the same concept but in terms of neurons , neuron as in human brain ? , again close enough, neuron but in Artificial Neural Network (ANN).

The activation function decides whether a neuron should be activated or not.

If you have seen an ANN, which I sincerely hope you do you have seen they are linear in nature, so to use non — linearity in them we use activation functions and generate output from input values fed into the network.

Activation functions can be divided into three types

1. Linear Activation Function
2. Binary Step Activation Function
3. Non — linear Activation Functions

Linear Activation Function

It is proportional to the output values, it just adds the weighted total to the output. It ranges from (-∞ to ∞).

Mathematically, the same can be written as

Implementation of the same in Keras is shown below,

Binary Step Activation Function

It has a specific threshold value that determines whether a neuron should be activated or not.

Mathematically, this is the equation of the function

Implementation of the same is not present in Keras so a custom function is made using TensorFlow as follows

Non — Linear Activation Functions

It allows ANN to adapt according to a variety of data and differentiate between the outputs. It allows the stacking of multiple layers since the output is a combination of input passed through multiple layers of the neural network.

Various non — linear activation functions are discussed below

Sigmoid Activation Function

This function accepts the input (number) and returns a number between 0 and 1. It is mainly used in binary classification as the output ranges between 0 and 1 e.g. you train a dog and cat classifier , regardless of how furry that dog is it classifies it as a dog not cat , there is no between , sigmoid is perfect for it.

Mathematically, the equation looks like this

Implementation of the same in Keras is shown below,

TanH Activation Function

This activation function maps the value into the range [ -1 , 1 ]. The output is zero centered , it helps in mapping the negative input values into strongly negative and zero values to absolute zero.

Mathematically, the equation looks like this

Implementation of the same in Keras is shown below

ReLU ( Rectified Linear Unit)

It is one of the most commonly used activation functions, it solved the problem of vanishing gradient as the maximum value of the function is one. The range of ReLU is [ 0 , ∞ ].

Mathematically, the equation looks like this

Implementation of the same in Keras is shown below,

Leaky ReLU

Upgraded version of ReLU like Covid variants .. sensitive topic …ok fine .. getting back to Leaky ReLU , it is upgraded as it solves the dying ReLU problem , as it has small positive slope in negative area.

Mathematically, the equation looks like this

Implementation in Keras is coming right below

SoftMax Activation Function

Its a combination of lets guess .. is it tanh , hmm not quite , ReLU ? no or its leaky counterpart .. mhh not quite …. ok lets reveal it .. it is a combination of many sigmoid. It determines relative probability.

In Multiclass classification , it is the most commonly used for the last layer of the classifier. It gives the probability of the current class with respect to others.

Mathematically, the equation looks like this

Implementation in Keras is given below

The whole notebook containing all codes used above

Google Colaboratory

If you wanna contact me , lets connect on LinkedIn link below

https://www.linkedin.com/in/tripathiadityaprakash/

---

Types of Activation Functions in Deep Learning explained with Keras was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

Loss functions are what make ANN (Artificial Neural Network) understand what is going wrong and how to get to that golden accuracy range, just like loss makes you cherish the profit and identify what went wrong.

Now coming to the mathematical definition of the loss function, it is used to measure the inconsistency between the predicted value (^y) and the actual label (y). It is a non-negative value, where the robustness of the model increases along with the decrease of the value of the loss function.

It’s straightforward actually, it’s basically how well your algorithm models the dataset, if its prediction is totally off, your loss function will output a higher number. If it is pretty good, it will output a lower number.

In this article, they are divided into three types

1. Classification Losses
2. Regression Losses
3. Model Specific Losses

Let’s discuss them in-depth,

Classification Losses

In this, there are again various categories,

Binary Cross Entropy Loss / Log Loss

This loss function is mainly used in the training of a binary classifier, it compares each of the predicted probabilities to the actual class output which can be either 0 or 1.

It then calculates the score that penalizes the probabilities based on the distance from the expected value.

The mathematics behind it is discussed below,

That means how close or far from the actual value.

Here, pi is the probability of class 1, and (1-pi) is the probability of class 0.

When the observation belongs to class 1 the first part of the formula becomes active and the second part vanishes and vice versa in the case observation’s actual class are 0.

Implementation in Keras is shown below

Hinge Loss

It is used for classification problems and an alternative to cross — entropy, being primarily developed for support vector machines (SVM), difference between the hinge loss and the cross entropy loss is that the former arises from trying to maximize the margin between our decision boundary and data points.

Mathematically the equation is shown below,

Implementation in Keras is below,

Categorical Cross Entropy

It is a SoftMax activation plus a Cross-Entropy loss. If we use this loss, we will train an image classifier using CNN to predict for each class present for example which bird class a bird image classifier belongs to .

It is used for multi-class classification.

Since activation functions are not discussed in this article and if you don’t know about it I i got you covered I have one more article covering activation functions in Deep Learning using Keras link to the same is below…

Types of Activation Functions in Deep Learning explained with Keras

Mathematically, the formula is shown below

Implementation in Keras is shown below,

Sparse Categorical Cross Entropy Loss

Sparse CCEL(Categorical Cross Entropy Loss) and CCEL both compute categorical cross — entropy, the only difference being how the targets/labels should be encoded.

When using Sparse Categorical Cross entropy the targets are represented by the index of the category (starting from 0).

Your outputs have a shape of 4×2, which means you have two categories. Therefore, the targets should be a 4 dimensional vector with entries that are either 0 or 1. This is in contrast to Categorical Cross entropy where the labels should be one-hot encoded.

Mathematically, the equation is

When to use which,

Implementation in Keras is shown below,

Kullback — Leibler Divergence Loss

Simply put, it is the measure of how a probability distribution differs from another probability distribution.

Mathematically, it can be explained using

KL divergence between two distributions P and Q.

The “||” operator indicates “divergence” or Ps divergence from Q.

KL divergence can be calculated as the negative sum of the probability of each event in P multiplied by the log of the likelihood of the event in Q over the probability of the event in P.

Implementation in Keras is as follow,

Regression Losses

In this, there are again various categories,

MSE ( Mean Square Error / Quadratic Loss / L2 Loss)

It is the average squared difference between observed and predicted values. e.g. in Linear Regression, we find lines that best describe the provided data points. Many lines can describe data points but which line to choose which best describes it can be found using MSE.

In below Diagram , predicted values are on the lines and actual values are shown by small circles.

Error is the distance between data point and fitted line.

Mathematically, it can be stated as

Implementation in Keras is shown below,

Mean Absolute Error / L1 Loss

Now we move on to Absolute values no messing around, all round up now!

It computes the mean of squares of errors between labeled data and predicted data. It calculates the absolute difference between the current output and the expected output divided by the number of output.

Unlike Mean Squared Error , its not sensitive towards outliers.

Mathematically , it can be written as

As seen in the formula modules is added giving only the absolute values for the error.

Implementation in Keras is shown below,

Huber Loss / Smooth Mean

The best of both of worlds, iOS & Android both phones together? Nah..

Its Huber Loss function , it combines MSE and MAE ,if loss value is greater than delta use MAE, if less use MSE.

Mathematically , it can be written as

Implementation in Keras is below

Now coming to Model Specific Losses they are mentioned with models in which they are used below,

MinMax (GAN Loss)

Generative Adversarial Networks uses this particular Loss and its a simple one ( that’s what they say).

Before this loss a little insight on GAN is needed so lets describe it in summary below

• Two Neural Networks are involved.
• One of the networks, the Generator, starts off with a random data distribution and tries to replicate a particular type of distribution.
• The other network, the Discriminator, through subsequent training, gets better at classifying a forged distribution from a real one.
• Both of these networks play a min-max game where one is trying to outsmart the other.

Now coming back to min-max loss, mathematically its equation is

The generator tries to minimize this function while the discriminator tries to maximize it and it saturates for the generator, making it stop training it is behind discriminator …sad generator noises….

This can be further categorized into two parts 1) Discriminator Loss and 2) Generator Loss

Discriminator Loss

Discriminator train by classifies both the real data and the fake data from the generator.

It penalizes itself for misclassifying a real instance as fake, or a fake instance (created by the generator) as real, by maximizing the below mentioned function.

• log(D(x)) refers to the probability that the generator is rightly classifying the real image,
• maximizing log(1-D(G(z))) would help it to correctly label the fake image that comes from the generator.

Generator Loss

While the generator is trained, it samples random noise and produces an output from that noise. The output then goes through the discriminator and gets classified as either “Real” or “Fake” based on the ability of the discriminator to tell one from the other.

The generator loss is then calculated from the discriminator’s classification — it gets rewarded if it successfully fools the discriminator, and gets penalized otherwise.

The following equation is minimized to training the generator,

This whole implementation of a Deep GAN Network along with loss in mentioned below as the loss is defined with respect to the model whole implementation needs to be referred.

docs/dcgan.ipynb at master · tensorflow/docs

Focal Loss (Object Detection)

It is extension of cross entropy loss function and it down-weight easy examples and focus training on hard negatives.

Focal loss applies a modulating term to the cross entropy loss in order to focus learning on hard misclassified examples. It is a dynamically scaled cross entropy loss, where the scaling factor decays to zero as confidence in the correct class increases.

Mathematics coming right up,

Focal Loss adds a factor (1−pt)^γ to the standard cross entropy criterion. Setting γ>0 reduces the relative loss for well-classified examples (pt>.5), putting more focus on hard, misclassified examples. Here there is tunable focusing parameter γ≥0.

Implementation in Keras is shown below,

The whole notebook with each losses implemented is mentioned below,

Google Colaboratory

References

https://towardsdatascience.com/understanding-the-3-most-common-loss-functions-for-machine-learning-regression-23e0ef3e14d3

https://medium.com/nerd-for-tech/what-loss-function-to-use-for-machine-learning-project-b5c5bd4a151e

https://neptune.ai/blog/gan-loss-functions

https://www.analyticsvidhya.com/blog/2020/08/a-beginners-guide-to-focal-loss-in-object-detection/

If you liked this article and want to read more like it , you can visit my medium profile and if you wish to connect here is my LinkedIn profile

https://www.linkedin.com/in/tripathiadityaprakash/

---

Types of Loss Functions in Deep Learning explained with Keras. was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

November 15–17 in Zoom Events & Decentraland

We have very exciting news!

The next Chatbot & Conversational AI Conference this November 15–17, 2022. This year’s event will be held on Zoom Events and in the Metaverse so that you can attend remotely from anywhere in the world!

The event will take you from Learning to Designing to Building a Conversation AI Agent during our 3-day event!

By the end of the event, you’ll have a Conversation AI Agent that you can share with your team or implement in your business, and you’ll get Certified!

Now let’s jump into the Agenda.

Zoom Events & the Metaverse

This year’s event will be held in ZOOM and Decentraland.

Why the Metaverse?

Conversational AI is already in the Metaverse. Inworld recently raised $50 Million for their platform, which creates AI-Powered Virtual Characters.

These characters can be used in games, in the metaverse, and online as brand ambassadors.

And yes, their co-founder, Kylan Gibbs, is also our Keynote Speaker!

Our goal was to give you a sneak peek experience of what will be like when Conversational AI, Web3.0, NFTs, and the Metaverse begin merging.

So, we created such an experience!

AI Bots in the Metaverses: This BOT will live in the Metaverse in Decentraland and attend our Conference. It solves one of the significant challenges of hosting events online: the social aspect of introducing people to each other.

This bot will use your Personality NFT to make high-quality intros!

Check out the Demo Here: https://play.decentraland.org/?position=5%2C118&

Killer Agenda

Our agenda is well thought out with you in mind. We start the day with the big picture of where the Conversational AI industry is now and where it is going.

The morning sessions feature Kylan Gibbs, co-founder & Chief Product Officer @ Inworld, who just raised $ 50 Million for creating AI-powered Digital Humans!

Morning Agenda: November 15, 2022

The Afternoon Sessions are where the rubber hits the road. We will do deep dives in areas of opportunity and look at execution.

The Afternoon has 3 Tracks.

• Industry Track: Sessions in Finance, Banking, HR, Hospitality, Customer Centers, &
• Execution Track: Design, Context, and NLU
• Live Podcast: Interviews and Q&A

Afternoon Agenda: November 15, 20

Certified Workshop in Conversational UX & NLU

Day 2: November 16, 2022 | Certified Conversational UX Workshop

CDI will do a live workshop via Zoom. In this workshop, you will learn how to design a Conversational AI agent by learning the most useful design frameworks.

This is a hands-on workshop where you will apply your learning to your project.

By the end of the workshop, you’ll be able to earn a Certificate.

Day 3: November 17, 2022 | Certified AI Workshop

On Day 3 we will give your bot life….

Botcopy will share its latest framework for creating a create NLU system that leverages data and eliminates waste. This is a dramatic shift in how NLU systems are currently built and instead leverages a lean and intelligent building approach.

So what is the approach? And what will you learn?

First, you will learn all the Dialogflow basics for creating a Conversational AI Agent, such as intents, entities, context, fall back loops, etc.

Second, you will learn the latest building framework which is a combination of very smart onboarding, fallback loops, and data gathering which creates a system for creating the correct FAQs in the proper manner without guesswork!

This is a hands-on, full-day workshop where you will apply put together this framework!

By the end of the workshop, you’ll be able to earn a Certificate.

Join us Today

If you haven’t reserved your spot, hit the link below and join us in November!

---

Chatbot Conference in the Metaverse was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

Building the Metaverse With Open Source — Bot Libre Metaverse

The virtual experiences that are promised from the metaverse will make several ordinary experiences more captivating and more inclusive as more persons have access regardless of distance and even time. Students, teachers, co-workers, family, friends, romantic partners and even strangers getting to know each other can find deep connection, productivity and success from the metaverse.

It’s clear that people already understand the immense value of web3, in a recent Forbes Digital Assets article, “the market was worth $478.7 million in 2020, and experts expect it to grow to $800 billion in just four more years.”

With such open enthusiasm, an open-source platform is the necessary engine that will drive the growth of the metaverse. By using open source to create the metaverse, it is possible to advance initiatives supporting the creation of decentralized, distributed, and interoperable virtual worlds.

Fortunately, as shared by opensource.com, all of this is accomplished utilizing just already-available, industry-standard web technologies, as well as open-source software and content. So, while each creator will customize their virtual world, the development concepts remain the same.

Through Bot Libre, you already have access to and can download all the tools and material you need to start building the metaverse. The Bot Libre platform enables influencers, gamers, and businesses to engage the Metaverse by integrating true artificial intelligence and chatbots. Bot Libre bots can interact with users and navigate 3D spaces, Bot Libre provides an extensive API, integrations, and SDKs for popular 3D platforms.

Bot Libre is making significant strides. Through our Mozilla Hubs integration users can create their own private 3D virtual spaces. These spaces can be used for events like conferences, virtual storefronts and online classes. By adding a Bot Libre chatbot to your 3D space, you will provide ease of access to all attendees and prevent any customer from being overlooked.

Also, Bot Libre’s OMNI AI model combines vision deep learning models with NLP deep learning models to provide multiple senses, integrated learning, awareness and navigation in the Metaverse. This will allow businesses to be functional and fully engaged with users in the Metaverse .With these models, users can explore a variety of different locations to chat with Bot Libre’s chatbots and bots can walk around different 3D spaces following or guiding users through the space. With Bot Libre’s AI algorithms, the chatbot’s avatars can avoid obstacles and find the shortest path to locations and objects.

Metaverse Solutions – Vision & NLP AI

BUT WAIT …It gets better

Bot Libre is now providing early access to unique metaverse solutions through our Beta Program.

We are looking for businesses, organizations, and developers that are interested in being early adopters of the Metaverse technologies, in order to shape, play, participate, and even profit from the Metaverse, by working with us to drive our open source metaverse AI solution. If you are interested in applying to be part of the program please contact sales@botlibre.biz .

Learned something? Please give us a

to say thanks and to help others find this article

---

Building the Metaverse With Open Source — Bot Libre Metaverse was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

Taking your business online is one of the most important tasks you can do, and there are E-commerce platforms out there like Shopify that make this task simple. Using Shopify, you can be up and running with your own E-commerce store in a matter of minutes.

Shopify lets you build beautiful, responsive, and professional-looking online stores without you actually having to know a lot about website development. And once you do have an online store, how do you ensure that you stay in touch with your customers 24 hours a day, 7 days a week? In one word: chatbots.

Adding a chatbot to your Shopify store can work wonders for your business, and, in this blog, we will teach you how you can go about adding a Dialogflow chatbot to your Shopify website.

Prerequisites:

• Shopify account that has access to third-party app integration, the first plan-Basic plan has access.
• Kommunicate account where you have already integrated the Dialogflow chatbot. If you have not integrated your chatbot yet, please refer to this video. Also, you can check out this tutorial for creating a Dialogflow chatbot.

Chatbot Integration using Dialogflow CX

1. Before Dialogflow CX integration with Kommunicate, first make sure you have created a project and Dialogflow CX Agent from the Dialogflow CX console.
2. Then you have to enable the Dialogflow API for your project and create a service account key. Please refer to the links below.

• Enable API
• Create Service Accounts

Create an agent

Build your Dialogflow CX chatbot

Build your Dialogflow CX chatbot as per your requirements. Once the chatbot building is completed, tap on the Agent Settings option to create a service account.

Create a Service Account

In the following step, you will be giving the ROLE to the service account, select Owner/Editor as a ROLE.

Once ROLE assignment is done, proceed to create a JSON key for Integration.

Create a JSON Key

Click on Manage Keys and Add a JSON key.

Once you click on CREATE, a JSON file download will start. Upload the service account key file in Kommunicate dashboard along with the AGENT ID.

To get the Agent Id:

Go to Dialogflow CX console >> Select the Project >> In the Agent you have created ‘Copy ID’

It will be in the following format, projects/test-covid-rwvr/locations/global/agents/e2c5d8a3-f416–4f32-bfc9-d986d540abd here the Agent Id is e2c5d8a3-f416–4f32-bfc9-d986d540abdb

Copy the Agent Id.

Enter the Agent Id in Kommunicate Dashboard and then “Save and proceed.”

On successful integration, the bot will be given an ID(botId) and will be listed under the Manage Bots section. The botId will be used to identify the bot in the Kommunicate system.

You can use a similar method for Chatbot integration with Dialogflow ES too.

Now that you know how to integrate a Dialogflow Chatbot, it is time to add the Kommunicate chatbot onto your Shopify website.

Integrate Kommunicate Chatbot onto your Shopify website:

Step 1: Log into your Shopify account and click on the Apps tab. Search for the Kommunicate app in the search text bar.

Step 2: After searching, you will see the Kommunicate app listed, click on it >> Add app.

Step 3: Click on the Install App button available in the top right corner.

After clicking on the Install App, it will redirect you to the Kommunicate dashboard. Log into the Kommunicate account to finish the integration.

Step 4: Once you finish the integration, click on Go to Store and check the Kommunicate widget and Save the changes.

Now that the chatbot installation is complete, visit your store to check your chat widget.

Do not forget to select your Dialogflow bot in the RULES section of the Kommunicate dashboard, so that you can enable the the same chatbot on your Shopify store.

Originally Published at https://www.kommunicate.io/on 5th Sep 2022.

---

How to Integrate Dialogflow CX Chatbot to Shopify was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

Baca dalam Bahasa Indonesia

I was doing sentiment analysis for a data journalism article for a bootcamp’s final project. The tweets are Indonesian tweets about Kominfo’s recent blocking of Steam (and some other overseas platforms). You can read my article here (it’s only in Indonesian though). By the way, I’m using Python on Google Colab.

At first, I was using TextBlob as explained here, but the result was bad. I was sure that at least for #BlokirKominfo tweets, it must be mostly negative. Yet, it was mostly neutral, or positive, depending on where you draw the lines (because TextBlob outputs polarity in range of -1.0 to +1.0). Well, TextBlob only supports English, so the tweets were translated into English. Also, TextBlob is rule-based, while state-of-the-art NLP uses neural network. But it’s not like I can just train my own neural network model in a few days. I’m a total noob at NLP. Luckily, Huggingface has a lot of pretrained model available for free, and there exists an Indonesian one.

So let’s get started.

1. Registering for a Twitter developer account

Twitter now provides an easy and free developer account registration. Of course, it will have a lot of limits. For example, the API calls you can do within 15 minutes is limited. You also can only query for the past 7 days. If you need more, you can upgrade your essential (basic) account by paying or by applying for an elevated or academic account. If you just need more API calls, go for the elevated one, since it’s still fairly easy. It also gives you access to V1 API. By default, you have access to V2 API and it’s still incomplete, for example, you can’t get trending data. If you need to go further than 7 days, go for the academic one, but it’s quite hard, since you need to have a legit research as a researcher or graduate student.

Okay, so how to get a developer account? Go to https://developer.twitter.com/ , click sign up at the upper right corner, then login with your Twitter account. You’ll then get a form. Fill the form, submit it, and you’re done. Make sure the save all of the tokens and secrets. But here we’ll only be using the bearer token.

2. Get tweets

First, you need to install Tweepy. You can do this using pip.

pip install tweepy

If you’re using google colab or jupyter, put a “!” at the front.

!pip install tweepy

Next, import it. “tw” here is just an alias to make it shorter.

import tweepy as tw

Now we create a client object using the bearer token. The wait_on_rate_limit argument being True tells Tweepy to wait for cooldown if you went past API rate limits.

client = tw.Client(
    bearer_token=bearer_token,
    wait_on_rate_limit=True
)

Then we can finally query tweets. We’ll be using search_recent_tweets. If you have academic account or paid account that supports past 7 days, you can use search_all_tweets instead.

result = client.search_recent_tweets(
    query,
    start_time=start_time,
    end_time=end_time,
    max_results=max_results,
    next_token=next_token,
    tweet_fields=tweet_fields
)

2.1. Arguments explanation

query would be your search query; it’s a required parameter. You can add “-is:retweet” at the end of your query to exclude retweets.

start_time and end_time defines the time window of your search. They’re in ISO 8601/RFC 3339 datetime format, which is YYYY-MM-DDTHH:mm:ssZ. It’s in UTC, so make sure you subtract your timezone difference to it. West Indonesia time is UTC+7, so today at 00:00 would be yesterday at 17:00 in UTC.

max_results is the max number of tweets in one API call. You can only have it 100 at most. If your query yields more than that, you’ll get a “next_token” attribute in the result’s metadata (not null). You can provide this next_token to get the “next page” of your query. If there’s no more tweets left to retrieve, next_token will be null. This should perfectly fit in a do-while loop, but python doesn’t have that, so let’s just set up a flag.

first = True
next_token = None
data = [] # This is just an example of combining the result
while first or next_token:
    result = client.search_recent_tweets(...)
    data = result.data
    next_token = result.meta["next_token"]
    # This is just an example of combining the result
    data += data
    ...

tweet_fields tells the API which tweet attributes/fields to retrieve. In Tweepy, it’s a list of strings or a comma-separated string of field names. For example, if you want to retrieve the like count etc, you’ll want the public_metrics field. In my project, I used these fields:

tweet_fields=[
    "author_id",
    "conversation_id",
    "created_at",
    "id",
    "lang",
    "public_metrics",
    "text"
]

Well, that’s all for the tweet search API.

3. Sentiment Analysis of Indonesian Tweets

Next, we’ll be using BERT to do sentiment analysis. We’ll use this model by mdhugol. It’s a sentiment classification model based on IndoBERT. It classifies text into positive, neutral, and negative ones (in this order of label). The website provides a field for you to try the model easily.

3.1. Init

Okay, for the model we’ll need to install transformer and emoji. Transformer package will be for the model, while emoji will be used to convert emojis into text that represents it. Just like before, add “!” at the front if you’re using google colab or jupyter.

pip install transformers emoji --upgrade

Next, we import them and define some constants. In addition to those packages, we’ll also use regex and html for cleaning.

from transformers import AutoTokenizer, AutoModelForSequenceClassification
from transformers import pipeline
import emoji
import re
import html

pretrained_id = "mdhugol/indonesia-bert-sentiment-classification"
label_id = {'LABEL_0': 'positive', 'LABEL_1': 'neutral', 'LABEL_2': 'negative'}

3.2. Text Cleaning

Before we do the sentiment analysis, we first need to clean the text. It’s not much, just removing or replacing some stuff. This function is based on preprocessing of indoBERTweet by indolem. It replaces user tags with @USER and urls with HTTPURL.

In addition to that, I’ll be replacing emojis with their text representation using emoji.demojize so the model will take emojis into account when determining the sentiment. However, it makes tweets longer, specially on those that spammed emojis. Meanwhile by default the model doesn’t support very long text. Twitter already has tweet length limit and it fits the model, but due to converting emojis, some tweets get very long. So I’m mitigating this by limiting repeating emojis to 3 repetition at most. I use regex to do this.

tweet = emoji.demojize(tweet).lower()
tweet = re.sub(r"#w+", "#HASHTAG", tweet)
tweet = re.sub(r"(:[^:]+:)1{2,}", r"111", tweet)
tweet = re.sub(r"(:[^:]+:)(:[^:]+:)(1+2+){2,}", r"121212", tweet)

Lastly, I unescape HTML-escaped characters using html.unescape. When using TextBlob, I found that some tweets uses escaped HTML characters. It made TextBlob throw an error. Not sure what it does to BERT, but I’ll just unescape them.

Anyway, here’s the full preprocessing functions. Just use the preprocess_tweet function.

def find_url(string):
    # with valid conditions for urls in string
    regex = r"(?i)b((?:https?://|wwwd{0,3}[.]|[a-z0-9.-]+[.][a-z]{2,4}/)(?:[^s()<>]+|(([^s()<>]+|(([^s()<>]+)))*))+(?:(([^s()<>]+|(([^s()<>]+)))*)|[^s`!()[]{};:'".,<>?«»“”‘’]))"
    url = re.findall(regex,string)
    return [x[0] for x in url]

def preprocess_tweet(tweet):
    tweet = emoji.demojize(tweet).lower()
    tweet = re.sub(r"#w+", "#HASHTAG", tweet)
    tweet = re.sub(r"(:[^:]+:)1{2,}", r"111", tweet)
    tweet = re.sub(r"(:[^:]+:)(:[^:]+:)(1+2+){2,}", r"121212", tweet)
    new_tweet = []
    for word in tweet.split():
        if word[0] == '@' or word == '[username]':
            new_tweet.append('@USER')
        elif find_url(word) != []:
            new_tweet.append('HTTPURL')
        elif word == 'httpurl' or word == '[url]':
            new_tweet.append('HTTPURL')
        else:
            new_tweet.append(word)
    tweet = ' '.join(new_tweet)
    tweet = html.unescape(tweet)
    return tweet

3.3. BERT model

Okay, now that the text are good, we can do the sentiment analysis. We create a tokenizer and a model using the name of the pretrained model (pretrained_id). It will download from Huggingface so you need internet connection. It’s quite big, but Google Colab runs on Google’s servers and uses Google’s fast connection. Next, we create sentiment-analysis pipeline using the model and tokenizer.

tokenizer = AutoTokenizer.from_pretrained(pretrained_id)
model = AutoModelForSequenceClassification.from_pretrained(pretrained_id)
sentiment_analysis = pipeline("sentiment-analysis", model=model, tokenizer=tokenizer)

Then, we just use the pipeline and convert the result label.

text = preprocess_tweet(text)
result = sentiment_analysis(text)
status = label_id[result[0]['label']]
score = result[0]['score']
print(f'Text: {text} | Label : {status} ({score * 100:.3f}%)')

That’s it. Really. Sentiment analysis done. Huggingface pretrained models made it very easy to do. Well, if you can find the right model that fits your use case, that is. Otherwise you’ll have to adjust it or even retrain the model.

4. Using the Proper Model for the Tweet Language

Despite the query being an Indonesian event, not all tweets are in Indonesian. We can use the “lang” field of the tweet to know what language it is in (according to Twitter). With that, we can do sentiment analysis of a tweet using the proper model for the tweet’s language. In my project, I used BERTsent by rabindralamsal for English tweets. However, do note that different models may use different labels. BERTsent, for example, also classifies tweets into positive, neutral, and negative, but in the opposite order. In the Indonesian model, LABEL_2 means negative, but in BERTsent, it means positive. Here’s the constants by the way:

pretrained_en = "rabindralamsal/BERTsent"
label_en = {'LABEL_0': 'negative', 'LABEL_1': 'neutral', 'LABEL_2': 'positive'}

Closing

Well, that’s about it. It took a while figuring out how to do it, but once you know how, it’s actually rather easy. But of course, I didn’t do much text preprocessing. I didn’t make the model myself either. Actually, I don’t even have NLP basics, so please correct me if I wrote something wrong. Or, maybe, if you want to add something, please just do so in the comments. Thanks for reading.

---

Indonesian Twitter Sentiment Analysis using Pretrained Neural Network Transformer (BERT) was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

WhatsApp For Business API 2022 — Complete Guide (Chapiter II)

Everything you need to know about WhatsApp chatbots in one place.

1 — Setup your WhatsApp For Business account

In this tutorial, we will use Twilio as our main provider for this use-case.
The biggest advantage using Twilio is the ability to perform tests using their shared WhatsApp For Business phone number. Once we create our account we will be able to join a channel and start messaging the chatbot.

1. Create your first Twilio account https://www.twilio.com/try-twilio

2. Go the console page and click on Create New Account

3. Let’s give a name to our account, I used : “Medium Bot Test”

4. Verify your email or phone number in order to continue the process

5. Answer these questions and make sure you use the same input

Tadaaa!! we have successfully created a free account with $15.50 credits.

6. Activate the sandbox by clicking on confirm

7. Send your first message to the bot

To begin testing, we will send a WhatsApp message from our device to +1 415 523 8886 with code join bow-gift.

8. Channel connection confirmation

2 — Generate your back-end application

The main component of this project is the back-end, it’s where all the processing occures. For this particular exercice, we will use JHipster to generate our back-end application in Spring Boot. It will speed up the development for a first MVP.

JHipster requirements :

• Node.js (LTS)
• Java 11 (JDK)

Make sure you have Node.js and Java 11 already downloaded and installed properly and then go to this tutorial to install JHipster : https://www.jhipster.tech/installation/

We will consider in this tutorial that you have already installed all requirements alongside with JHipster.

Once you launch Node.js command line, create a folder called MediumBotBundle or whatever name that suits you, after that, run this command to start JHipster :

cmd -> jhipster

After that, the generator will ask you about the type of application, select Monolithic application.

After few minutes, the application will be generated and you will be able to start development. I personally use Intellij Community Edition as my main IDE and I highly recommand it for your future projects.

Create a class named “WhatsAppBusinessController” in web.rest package. The purpose of this class is to intercept all requests sent by Twilio to our WebHook when a user communicates with the bot.

Below is the JSON received in our interceptor :

We have 3 main attributes from this JSON message :

• From : The phone number of the sender
• To : The phone number of the receiver
• Body : the content of the message

To be continued..

---

WhatsApp For Business API 2022 — Complete Guide (Chapiter II) was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.

Bad UI Patterns That Hurt Your Business.

1) SHOW USER UNAVAILABLE OPTIONS

Users shouldn’t be guessing which option is available or not when they are searching for something.

Take this search as an example. National Car Rental website shows the user all the locations available to book a car for the dates already selected, but when you try to select one of them, two errors appears: either the store doesn’t open at the select date and time or the store is not valid for the promotion code added.

As you can see, there are many stories in the São Paulo (City in Brazil) area that supposedly is “available” to book (the red error only shows after your try to book), but after 2 minutes of clicking on all of them, only 2 or 3 were actually accepting bookings at given filters.

The only exception I’d like to mention is when the unavailable option is placed in the results as a marketing strategy.

As an excellent example, I can show Booking.com — which by the way does an awesome job with their UX.

They do show unavailable options (but notice here that the message about it is already visible to the user without having to click, it doesn’t leave the user confused thinking that’s available), but to induce a sense of FOMO, i.e., fear of missing out. They are indirectly saying “hey, look, this hotel is sold out because you were late, act now or you’ll miss more opportunities”.

SOLUTION

The user is already doing a search, which basically is a filtering action, so give it the proper treatment and only show the available results.

2) TEXT HARD TO READ BECAUSE OF THE FONT WEIGHT AND COLOR

SOLUTION

Never use fonts too light in small texts, and always check with web accessibility evaluation tools if your design is compliant to WCAG rules to avoid law suits

3) FORGET TO DISABLE IT AND LACK OF FEEDBACK AFTER CLICKING ON A BUTTON.

Sometimes we need to load content after the whole site is already there. But if that new content pushes everything down when being placed, you got a UX problem here.

Since the interface is already loaded, the user might be in the middle of an action, a click, drag, and drop. And having the content pushed down might cause unexpected behavior if the user clicks somewhere at the same moment that the new content appears.

The example below shows a shopping cart after adding a product. The e-commerce then loads related products on top of the cart after the content has already been loaded. If the user tries to change the quantity of the first product in the cart right at the moment that the related products are being loaded, it ends up clicking on one of them and going to the product page, frustrating the experience of buying that first product (Yeah, I did that many times, I always forget to wait for the related products, but this shouldn’t be an issue the user needs to deal with).

---

Bad UI Patterns That Hurt Your Business was originally published in Chatbots Life on Medium, where people are continuing the conversation by highlighting and responding to this story.