Scribd
Upload
61 views2 pages

Research

The document discusses research done at Vize on improving various aspects of their application and database architecture. It covers: 1) Separating calculations from the portal to improve scalability and performance. 2) Developing metrics for measuring collaboration including leadership, sentiment, learnership, and connectivity scores. 3) Redesigning the Vize system to move calculations to dedicated services and introduce roles and subscriptions. 4) Normalizing the database to third normal form to improve structure and query performance.

Uploaded by

Extinct Reviews
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
61 views2 pages

Research

The document discusses research done at Vize on improving various aspects of their application and database architecture. It covers: 1) Separating calculations from the portal to improve scalability and performance. 2) Developing metrics for measuring collaboration including leadership, sentiment, learnership, and connectivity scores. 3) Redesigning the Vize system to move calculations to dedicated services and introduce roles and subscriptions. 4) Normalizing the database to third normal form to improve structure and query performance.

Uploaded by

Extinct Reviews
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

Vize – Research and Innovation Document

Scalability Research
Problem: How can we make the application scalable. Current application is slowed down when
multiple users are on the website. Currently the AI is running on the same server as the
integration data pulling.
Solutions:
 Separate AI calculation and data pulling. Each integration is its own application.
Scores for each integration is also calculated using separate applications.
 Utilize async to use multiple threads of the CPU to perform calculations
simultaneously.
 Utilize a daily scores application to handle the calculations on a separate server
which will no longer slow down the portal.
 Portal application now has zero calculations – only function is to handle
authentication to the website. Metrics are now pulled from the scores application.

Collaboration Metrics Research


Problem: How do we define the metrics our application calculates and how do we develop a
formula for said metrics. Current metrics include, Overall Health Score, Sentiment score,
Leadership and Learnership scores, Peer connectivity, Organization connectivity, Comfort score.
Solutions:
 Leadership score – A percentage of the number of questions users are asking this
individual compared to all other conversations this person is having. If a user is
being asked many questions Vize considers them a leader.
 Sentiment score – A score developed to determined how positive the messages
sent by an individual are. This can be used to extrapolate how happy or unhappy a
user is with their job.
 Learnership score – A percentage of the number of questions this user is asking
compared to all other conversations. If a person is asking many questions Vize
considers them to have a high learner score.
 Peer Connectivity – Developed a score to reflect how comfortable a user feels
around their peers. This takes into consideration how positive the conversations a
user is having with other members of the team as well as investigates if the user is
speaking with all the members of their assigned team.
 Organization Connectivity – Developed a score to reflect what the user feels
about their organization. This score takes into consideration the fraction of
suggestions and complaints the user is making compared to the total messages
sent by the user. This is used to determine if the user is complaining about the
organization and if they are making suggestions regarding how the organization
operates.
 Overall Health score - A

Vize System Redesign Research


Problem: Current Vize system is slow. All calculations were handled by the front end portal
which made the website slow when multiple users were online
Solutions:
 Created a ERD diagram to reflect a new Vize system which included a reduced number
of bi-directional relationships.
 Converted previous “department” design to a simpler organization design which
removed “personal department” features.
 New “organization” entity can have an owner as well as multiple admins.
 Roles introduced to limit what users can view on the site. Roles for clients include
owner, admin, member roles.
 Created “Super Admin” for Vize employees, this is used to onboard new clients.
 Created a “subscription” entity used to determine what subscriptions the current client
has paid for and limit the websites functionality based on the subscriptions. This
includes, collaboration, social, and search subscriptions.

Vize Database Redesign Research


Problem: Previous Vize database did not follow 3NF form. Database is confusing with multiple
schemas and tables that do not properly reflect their purpose and the database is performing
poorly with query speed.
Solutions:
 Researched 1NF, 2NF and 3NF rules and implemented them into the new database
structure.
 Modelled the database based on the ERD diagram developed during the Vize system
redesign.
 Separated data into multiple schemas including separate schemas for portal, each
integration as well as the scores

You might also like