Artificial Intelligence in Customer Service

“Ok… But how?” If you’re thinking that this is all really nice but you still don’t get it, don’t worry, we know. This is why we decided to dedicate an entire page to it.

Artificial Intelligence in Customer Service

“Ok… But how?” If you’re thinking that this is all really nice but you still don’t get it, don’t worry, we know. This is why we decided to dedicate an entire page to it.

Take the tour

How our A.I. works

A magician never reveals his secrets… but we’re not magicians, just really good at customer service. Our solutions work with artificial intelligence to understand every question and message and answer your customers in an efficient and natural way, without wasting their time. Take the tour!

How our A.I. works

A magician never reveals his secrets… but we’re not magicians, just really good at customer service. Our solutions work with artificial intelligence to understand every question and message and answer your customers in an efficient and natural way, without wasting their time. Take the tour!

1. Normalization (NPL)

  • Grammatical correction
  • Automated spell-checking
  • Verification of the existence of meaning
  • Elimination of unnecessary characters
Message: did you opn a new plce at NY?
  • opn = opened
  • plce = place/branch
  • NY = New York

Normalization: Did you opened a new branch at New York?

2. Classification

  • Search and categorization of meaning
  • Identification of relevant actions and products
Message: did you opn a new plce at NY?
oppnd = action
plce at NY = location

3. Contextualization

  • Short term memory
  • Channel
  • Conditions
  • Flows (Nested questions)
Message: where is it?

Context:

  • WEB: www.phonelife.mx
  • Country: USA
  • User: Not logged-in

Memory:

Previously mentioned location: new New York branch


Answer: It’s located at 742 Evergreen Terrace

4. Intention identification

  • Entity identification and intention relevance
  • Word’s distance
  • Personality intention checking
Message: where is it?

Ways to ask:

  • Where is the new New York branch?
  • Can you tell me the address of the new New York branch?
  • Location of the new New York Branch.

Identified Intention: Location of the new New York Branch.

5. Semantic Assistance

  • Deep Learning based intention identification.
  • Detection of the intention regardless of the written text.
  • Less need for “ways to ask”.
Message: when do you close?

Ways to ask:

  • What is the attention hour of the new New York branch?
  • What is the attention time of the new New York branch?
  • What is the opening hours of the new New York branch?

Identified Intention: Attention time of the new New York branch.

6. Selection of the best response

  • Precision evaluation.
  • Disambiguation.
  • Suggestions for questions that could not be solved or with little assertiveness.
  • Sending to continuous improvement intentions not found.
Message: and in San Francisco?

Intention not found

Suggested intentions:

  • Existence of stores in San Francisco
  • Store hours in San Francisco
  • Location of branches in San Francisco

7. Empathic response

  • Make sure your user is comfortable!
  • Complements (assisted navigation, carrusel, faqs, images, videos, etc.).
  • Integrations.
  • Live chat transfer.

Message for the user

We would love you to visit us! Here’s a map of our stores so you can find the nearest to your home.
MAP COMPLEMENT

1. Normalization (NPL)

  • Grammatical correction
  • Automated spell-checking
  • Verification of the existence of meaning
  • Elimination of unnecessary characters
Message: did you opn a new plce at NY?
  • opn = opened
  • plce = place/branch
  • NY = New York

Normalization: Did you opened a new branch at New York?

2. Classification

  • Search and categorization of meaning
  • Identification of relevant actions and products
Message: did you opn a new plce at NY?
oppnd = action
plce at NY = location

3. Contextualization

  • Short term memory
  • Channel
  • Conditions
  • Flows (Nested questions)
Message: where is it?

Context:

  • WEB: www.phonelife.mx
  • Country: USA
  • User: Not logged-in

Memory:

Previously mentioned location: new New York branch


Answer: It’s located at 742 Evergreen Terrace

4. Intention identification

  • Entity identification and intention relevance
  • Word’s distance
  • Personality intention checking
Message: where is it?

Ways to ask:

  • Where is the new New York branch?
  • Can you tell me the address of the new New York branch?
  • Location of the new New York Branch.

Identified Intention: Location of the new New York Branch.

5. Semantic Assistance

  • Deep Learning based intention identification.
  • Detection of the intention regardless of the written text.
  • Less need for “ways to ask”.
Message: when do you close?

Ways to ask:

  • What is the attention hour of the new New York branch?
  • What is the attention time of the new New York branch?
  • What is the opening hours of the new New York branch?

Identified Intention: Attention time of the new New York branch.

6. Selection of the best response

  • Precision evaluation.
  • Disambiguation.
  • Suggestions for questions that could not be solved or with little assertiveness.
  • Sending to continuous improvement intentions not found.
Message: and in San Francisco?

Intention not found

Suggested intentions:

  • Existence of stores in San Francisco
  • Store hours in San Francisco
  • Location of branches in San Francisco

7. Empathic response

  • Make sure your user is comfortable!
  • Complements (assisted navigation, carrusel, faqs, images, videos, etc.).
  • Integrations.
  • Live chat transfer.

Message for the user

We would love you to visit us! Here’s a map of our stores so you can find the nearest to your home.
MAP COMPLEMENT

Start changing the future of your conversations

Start changing the future of your conversations

Don’t take our word for it.

Since 2012, we’ve been partnering up with companies that share our vision.

READ THEIR STORIES