Email Automation_
_Tidy Up Customer Service
Your company can do more with the same resources - just let AI do the robot work so that your agents can take care of customers!
Explore benefits Contact us
Dispatching
Quality Assessment
Attachment Processing
Too many inbound emails? Too much effort spent on tagging or sorting them into specialised inboxes?
Too limited view of service quality? Too little control over agent performance?
Too many attachments? Too much time spent on opening & processing them?
Dispatching
Too many inbound emails? Too much effort spent on tagging or sorting them into specialised inboxes?
Learn more
Quality Assessment
Too limited view of service quality? Too little control over agent performance?
Learn more
Attachment Processing
Too many attachments? Too much time spent on opening & processing them?
Learn more
Email Automation:
Dispatching_
Objectives
Relieving your agents from repetitive work by automating the process of dispatching emails.
Benefits
Save time and manpower through near-human precision & perfect adaptation. SentiSquare AI self-learns to understand the meaning and context of your emails and your inboxes.
Our Solution
Wired into your processes through API, SentiSquare AI labels customer emails in real time as they come. Deployment takes only 14 days.
Success story:
Dispatching at E.ON_
Case
50,000+
emails per month;
time to dispatch
Before
2 min 39 sec
per email + backlog
After
0.3 sec
per email
NO backlog
50,000+ emails / month to info@eon.cz
280 specialised inboxes
14 FTE needed for manual dispatching
E.ON’s Challenge
On average, human agents needed almost 3 minutes to dispatch one email to the correct inbox. But the amount of incoming queries was rising and the email backlog was getting bigger and bigger. E.ON was searching for a solution to speed up the painful, time-consuming process.
SentiSquare’s Solution
SentiSquare deployed self-learning, natural language processing (NLP)-focused AI. By learning to understand the meaning and context of E.ON’s past emails, SentiSquare AI reached near-human accuracy. Wired into the process through API, SentiSquare AI started to label customer emails in real time as they came.
E.ON’s Results with Sentisquare
Since the deployment of SentiSquare AI, emails are getting through to the right agent immediately. Through continuous learning, SentiSquare AI is getting even better over time. As a result, E.ON is saving time and effort – letting humans connect and add real value instead of doing robot work!
“SentiSquare AI understands meaning of text and that is what you need if you are drowning in texts”
Jana Krafková, Customer Service Center @ E.ON
Email Automation:
Quality Assessment_
Objectives
Gaining real-time overview of customer service quality through automatic checking of outbound emails.
Benefits
Achieve superior quality assurance through 100% coverage & consistent evaluation. SentiSquare AI self-learns to check all outbound emails emails for grammar, empathy, formatting, & proper greeting.
Our Solution
Based on your past emails, SentiSquare AI adapts semantic models to your operations and checks outbound emails via API as agents write them.
Success story:
Quality Assessment at Kiwi.com_
Case
80,000+
emails per month;
% QA coverage
Before
1 %
After
100 %
80,000+ emails / month
QA team randomly checking 1% of agent-written emails
No way to know which individual agents to focus on; overall performance could not be monitored, either
KIWI.com’s Challenge
On average, the Quality Assessment team only managed to cover 1% of all the emails sent by KIWI.com’s agents. Emails were being checked on nearly random basis – the team had no way to know which individual agents to focus on. Overall performance could not be monitored, either.
SentiSquare’s Solution
SentiSquare deployed self-learning, natural language processing (NLP)-focused AI to evaluate emails in terms of grammar, empathy, format, & proper greeting. Wired into the process through API, SentiSquare AI started to assess agent-written emails in real time as they went out.
KIWI.com’s Results with Sentisquare
Since the deployment of SentiSquare AI, 100% of agent-written emails are getting checked immediately. Through continuous learning, SentiSquare AI is getting even better over time. Managers at KIWI.com now know precisely how the quality of customer service is looking & which agents need more training!
Email Automation:
Attachment Processing_
Objectives
Recognizing specific types of attachments through a combination of image recognition and natural language processing (NLP). Automated editing of document scans & images them into a correct, unified format.
Benefits
Speed up query resolution. Make email attachments easier to handle for your agents by recognising & extracting specific types of attachments.
Our Solution
Using an API, SentiSquare AI will recognise and preprocesses in real time all email attachments which you define as important for your operations.
Success story:
Attachment Processing at E.ON_
Case
25,000+
attachments per month
50+
types of form
Before
Manually processing
After
Automatically dispatched
25,000+ email attachments / month
Need to recognize priority attachments
Manual information extraction
E.ON’s Challenge
On average, human agents needed almost 3 minutes to dispatch one email to the correct inbox. But the amount of incoming queries was rising and the email backlog was getting bigger and bigger. E.ON was searching for a solution to speed up the painful, time-consuming process.
SentiSquare’s Solution
To classify and extract attachments, SentiSquare deployed a combination of optical character recognition (OCR) and natural language processing (NLP). Through machine learning, SentiSquare AI reached near-human accuracy. Wired into the process through API, SentiSquare processes emails in real time.
E.ON’s Results with Sentisquare
Since the deployment of SentiSquare AI, emails are getting through to the right agent immediately. Through continuous learning, SentiSquare AI is getting even better over time. As a result, E.ON is saving time and effort – letting humans connect and add real value instead of doing robot work!