Transcript:
0:00 [Music]
0:04 Welcome to this North52 tutorial. Today
0:07 we'll explore how to automatically
0:08 process emails using Microsoft database
0:11 AI functions integrated with the North
0:13 52 business rules engine. By the end of
0:15 this video, you'll understand how to use
0:17 AI to classify, analyze sentiment,
0:20 summarize, and automatically reply to
0:22 emails in Dynamics 365.
0:24 Processing customer emails manually can
0:26 be timeconuming and inconsistent. But
0:28 what if we could leverage AI to
0:30 automatically classify emails, determine
0:32 their sentiment, create summaries, and
0:34 even draft replies? We'll be focusing on
0:37 four powerful data versse AI functions.
0:40 AI classify to categorize emails as
0:43 problems, questions or requests. AI
0:46 sentiment to determine if the tone is
0:48 positive, neutral, or negative. AI
0:50 summarize to create concise summaries.
0:53 And AI reply to generate draft
0:55 responses.
0:57 Our solution consists of three key
0:59 business rules that work together
1:00 seamlessly. When an email enters our AI
1:03 processing queue, we strip the HTML from
1:05 the email body to prepare clean text for
1:08 AI analysis. We analyze the email using
1:11 three data versse AI functions and
1:13 create a case with intelligent
1:15 classification and priority assignment.
1:19 We generate a draft email reply using AI
1:22 reply and link it to the newly created
1:24 case. Let's dive into the technical
1:26 setup. We'll create a North52 formula
1:29 that triggers when items are added to
1:31 our Q items table.
1:33 We'll set the pipeline stage to post
1:36 operation asynchronous for optimal
1:38 performance. Our solution uses three
1:40 decision table sheets. Let's examine
1:42 each one. The validation sheet ensures
1:45 we only process emails assigned to our
1:47 AI processing queue. If conditions
1:49 aren't met, processing stops here.
1:53 The analyze email and create case sheet
1:56 is where the magic happens. Here we see
1:58 three AI function calls and action.
2:02 The AI classify function uses this
2:04 syntax to categorize emails into
2:06 problem, question or request types.
2:09 Notice how we pass the email text and
2:11 define our classification categories.
2:16 AI sentiment analyzes the emotional
2:18 tone, helping us set appropriate case
2:20 priorities automatically.
2:23 AI summarize creates the concise
2:25 descriptions that capture the essence of
2:27 lengthy customer emails. Finally, our
2:31 create draft email reply sheet uses AI
2:33 reply to generate professional
2:35 contextual responses that human agents
2:37 can review and send. Let's see our
2:40 solution in action with some real
2:42 examples. Here's a frustrated customer
2:45 email expressing dissatisfaction with
2:47 the service. Watch how AI processes
2:50 this. Our system correctly identified
2:53 this as a negative sentiment, classified
2:55 it as a problem, created a high priority
2:57 case, and generated an appropriate
2:59 summary.
3:02 Notice how the professional, empathetic
3:04 draft reply that addresses the
3:06 customer's concerns while maintaining a
3:08 helpful tone.
3:12 Now let's process a positive customer
3:14 email to see how AI categorizes this.
3:18 The positive sentiment is also resulting
3:20 in a high priority. This shows how our
3:23 AI system prioritizes both negative
3:26 feedback requiring immediate attention
3:28 and positive feedback that represents
3:30 engagement opportunities.
3:32 Notice how the AI generator reply for
3:34 the positive email has a completely
3:36 different tone. It's appreciative and
3:38 encouraging, matching the customer's
3:40 positive sentiment.
3:43 This North52 solution delivers
3:44 significant benefits, faster response
3:47 times, consistent classification,
3:49 appropriate prioritization, and reduced
3:51 manual effort. For more details about
3:54 data versse AI functions, check out the
3:56 Microsoft Learn documentation linked in
3:59 the support article.
4:01 Thanks for watching. Visit
4:03 support.n2.com nor2.com for more
4:05 automation solutions and examples from
4:08 our knowledge base. Don't forget to
4:10 subscribe to our YouTube channel.