Real-time Bidirectional Mail Solution

Real-time Bidirectional Mail Solution. Solution Design.

BDE. Problem statement. 800+ Mail Boxes Not Real-time No accurate categorization SLA: 8 Hr. No workflow – manual selection No Escalation unless set up No Duplicate control No-reply email addresses are received.

BDE will provide the following:. Move email to “real Time”.

[Audio] The solution is broken up into 7 key points – Receiving emails from +800 Mailboxes ( Some would need to be decommissioned as we receive small volumes) – Average request volume per Month is 500K Emails metadata is extracted from all the mailboxes and sent to an SQL database The metadata flows to the ReadMe engine to validate the requestor i.e., Existing Member, Scheme, Unknown Once data is validated , each email that has a member validated will run through the Nexus Database to get additional information (This will help our digital workers process the request faster and eliminate the need to extract data from the front end to validate) All the emails will be sent to common mailbox (This is the Digital Workers mailbox). The mailbox has an Azure Natural Language Processing model attached to it. The model will scan the emails and split the emails into Sentiment buckets within the Mailbox i.e., Claims , Claims without Details, Tax certificate, Membership etc. The digital will use each Sentiment bucket as a Task queue i.e. Go into a bucket Acknowledge the request to the customer Understand which category the request belongs too Straight Through processing , Request more information, Refer to team. Digital worker will process the request in Nexus to get the information for response back to customer Respond to customer with the request and close the query ID in Nexus In order to maintain the solution, we will include Logic apps ( A3) to communicate with the relevant teams should there be system issues experienced.

Membership. Finance. Claims. Departments.

First roll out processes. Card Request Members requesting new membership cards.

Fedhealth. Bonitas. Polmed. Schemes.

1. E- mail received from member. E-mail received from member.

Agent referred. 6 E-mail follow-ups. 4 Days to resolve.

[Audio] The above graph depicts the volumes of the queries in the workgroups we worked with Blue – readme volumes – (speak to undelivered, spam) Actual Yellow - STP – reflected Grey - Drivers reflected Why look at orange bar (projected ROI for rollout purposes 150 000) Confirm Contributions CSAT50[ break]% Cards: CSAT 58[break]% Claims: 20.5[break]%.

5899 Bonitas Of the 87636 e-mails handled by the BU 22707 related to Membership/Finance/Claims.

Arrears confirmation.

Team Mate A. Team Mate B. Your Text Here You can Add Some Brief Text That will Explain Your Tittle.

Step 1. HEADING Here Add Some Own Text of Yours. Step 2.

38. Project A. PERCENT. TEXT HERE. Project B. 62.

38. Project A. PERCENT. TEXT HERE. Project B. Project C.

Add Your Text Here You can Add Some Brief Text That will Explain Your Tittle.

Add Your Text. Here You can Add Some Brief Text That will Explain Your Tittle.

Add Your Text. Here You can Add Some Brief Text That will Explain Your Tittle.

BDE. Thanks for listening to us, now you can ask your questions.

Timeline. 4 days to resolve. 6 email follow ups. 3 minutes.

Timeline. 8 days to send acknowledgment. No follow ups.

[Audio] Use cases. Timeline. 6 days to resolve. 2 live chat follow ups.

[Audio] 12 Bots depending on the email distribution curve and peaks 5 min within a max of 1 hour Note: research benchmark for CSAT scores.