Theme - Finance - Accounts Payable Invoice Process Automation using IQBOT integrated with AARI virtual assistant

Inspiration - Invoice processing is a mundane repetitive task performed by Accounts Payable team across the Industry. Also team spends more time for providing Invoice status to Vendor(PROD Support) which is labor intensive

What it does - Automating End to End Invoice Process for Accounts Payable Department. Also integrate with AARI for PROD support to reduce the time being spent bu Business

How we built it -

  1. Created Project Workflow design
  2. Database Design
  3. Created IQBOT Learning Instance and trained Invoices with ABBYY engine
  4. Created re-usable Python code for data processing during IQBOT training(Ex: converting Date to required format)
  5. Created Modular task bot for Email attachments download, IQBOT upload & download. Merge csv file data to Access database
  6. Created AARI web form and built task bot to fetch Invoice details for Human verification

Challenges we ran into

  1. Unable to create hyper link in Gmail notifications. Workaround: Created file path as text instead of hyperlink
  2. Ran into Database insertion error due to Data type mismatch which was solved after research
  3. File copy & delete intermittent errors. Workaround: Provided delay and condition to check if file size is same in source & destination folder

Accomplishments that we're proud of

We have come up with complete SDLC process, so it can benefit anyone seeing our Project and can implement the project with ease with all different templates provided in submission

  1. Detailed design for Project process
  2. Detailed database design
  3. Modular task for re-usability
  4. Credential vault implementation to avoid changes of critical variables
  5. Unit testing of all the modules
  6. Creation of Dashboards using Power BI
  7. Created IQBOT LI dashboard as IQBOT Control room doesn't have Drill down functionality to see the trend
  8. Creation of python re-usable functions for IQBOT data processing

What we learned

  1. Usage of A360 platform as we were mostly on AA v11 for last couple of years
  2. Bot games helped us to learn different packages and connect with Dev community and learn different approaches to solve problems by increasing the efficiency
  3. IQBOT success tips

What's next for IPABOT

  1. To scale up the project across different departments
  2. To process other unstructured documents and effectively utilize IQBOT capability
  3. To provide the document view directly in AARI web form for Business users

Built With

  • aari
  • access
  • api
  • database
  • framework
  • gmail
  • iqbot
  • modular
  • powerbi
  • python
Share this project: