2. SWIFTSENSE
• Supervisor :
• Mr. Hammad Mehmood
• Co-Supervisor :
• Dr. Ghulam Abbas
• Group member :
• Affan Javaid
• Abdullah Khan
• Waleed Khan
3. BRIEF OVERVIEW
• SwiftSense - It is a Sale Forecasting system.
• Swift = quick , Sense = observation
• It means how our system will quickly predict the sales of a franchise company by observing its previous
and current data.
• Ease sales operations/forecasting in franchise companies
4. SWIFTSENSE
Flaws in Franchise Companies System
• They are using an obsolete system in their sales department.
• They rely more on data-storage and excel-sheets.
• Lack of communication between departments and offices.
• Less performance
5. PROBLEMS IN EXISTING SYSTEM
• Local stores still use Pen/Paper System.
• Individual systems for every department.
• Delay in communication/approval.
• Example : Comparison between Walmart/Zara and Bata .
6. PROPOSED SYSTEM - SWIFTSENSE
• Develop an Integrated System – overcome delay in communication/approval.
• Central Database
• Use AI in Sales Forecasting . To increase the performance , efficiency and revenue of a company.
• Apply Supply chain concepts- Just In Time (JIT) , Point of Sale , Available to Promise (ATP) etc .
7. FUNCTIONAL & NON-FUNCTIONAL REQUIREMENTS
Functional Req Non functional Req
Central DB Storage Safety
Store Management Security
Sales Forecasting Software Quality
Point of Sale / Just in Time
Performance measurement
Super Admin
8. SYSTEM UNIQUENESS
• It will be the first system in the history of Pakistan in which will provide all these major services in a
single system . This makes our system unique
• It will be an integrated system . All major departments like Marketing , Admin , Manufacturer etc will be
integrated.
• It will provide Sales forecasting using AI
• It will implemented on major supply chain concepts
9. GOAL
• There are 3 main goals of our system
• To replace existing system used by franchises
• To increase the performance and efficiency of the franchise company with latest technologies and concepts
• To maximize their revenue/profit.
10. SCOPE
• We will target the local market / Franchises.
• According to report published by Karachi Chamber of Commerce and Industry,
• There are more than 160 Franchise companies in Pakistan.
• 4-5 Franchise companies are registered every year
• According to report published by Forbes (w.r.t Falcon Fatemi).
• There is 7-13% increase in revenue of a franchise by integrating AI in their Sales Forecasting
• If our system can even increase 3-5% of the Franchise’s revenue. It would be profitable for
both company and us
11. AI – SALES FORECASTING
• Why we need AI in sales forecasting?
• Forecasting is essential to how businesses run.
• It enables a company to set a budget, make headcount decisions, fund marketing and product
investments.
• It is an important part of supply chain.
• With AI , Sales Forecasting prediction efficiency increases & it is less prone to errors etc.
12. AI – SALES FORECASTING
• How our AI in sales forecasting will work?
• Choose a particular AI algorithm (yet to be finalized) - which fulfills our requirements/needs.
• Train our AI on different Data Sets using that selected algorithm.
• Implement Trained AI on our system.
• Predict the next month sales.
• Compare it with the verified result
• Unit testing – catch & minimize errors as much as possible
13. AI – SALES FORECASTING
• For Training of our AI . We will gather data(different data sets) from
• Kaggle
• Data.world
14. PROPOSED TECHNOLOGIES – DEVELOP SYSTEM
• System (SwiftSense) will be a website.
• Frontend : React JS
• Backend : Node JS
• Database: Mongo DB , Firebase
• AI : Node JS
• Deployment : Azure , Docker etc (not finalized yet)
18. ARCHITECTURAL DESIGN
• We will use MVVM (updated form of MVC) architectural design.
• Why MVVM:
• The View — UI layer that users interact with,
• The ViewController —has access to the ViewModel and handles user input,
• The ViewModel — has access to the Model and handles business logic,
• The Model — application data source
19. MVVM
• Views :
• Views are UI which are in form of React component. Every view has its different react
component
• Models :
• We will use cloud Mongo DB as database. All of our data will store in database. Every user has
access to different data
• View-Model :
• View Model controls the view from models. We will use NodeJs to control this.
• Admin can check the total sales of each store.
• Manager will calculate the daily sales and send it to admin