Workflow automation

Python Flask framework based web application for workflow automation, increase of business processes performance, granting operational transparency and making people happier (2018-2019)

Idea comes from the customer needs to make a solution for business performance boosting. Audit of business as a kickoff followed by business process optimization supported by implementation of suitable workflow automation system. Not a full ERP solution, just simple and nice workflow automation. There are many solutions on the market, some of them a "no-code", some are complex and come close to ERP systems. Lightweight tool granting stability and ease of the support, CI/CD processes with acceptable TCO (Total Cost of Ownership) was very hard to find. As a result - created new solution suitable for business.


Audit. To solve the problems first one needs to find the weak spots. Project started from heliview of the business operations with detailed description of business processes in BPMN notation. Some crucial bottlenecks and weak spots were found during audit. Main business problem was identified as piecewise automation of end-to-end main business process with bulky information systems. This led to increased amount of manual labor, thus increased costs. TCO of existing information systems was pretty high compared to the added value to services and products.

Business process. After building as-is business processes maps project continued with processes to-be models. Conclusion was indefinite, none of existing IT solutions could remove the main problem of end-to-end process - piecewise automation. Building another layer of IT infrastructure like ESB (Enterprise Service Bus) for connecting all the systems was a very bad idea. Another approach was selected, make a pinpoint business process reengineering with small changes of working instructions for personnel supported with a new lightweight software solution.

Custom software solution. Creating of architecture of future information system showed that the most boosting result with low TCO can be achieved using ready software frameworks. Python and Flask were chosen due to simplicity, modularity and vast amount of prebuilt software libraries. General technical requirements were created for a team of developer. Detailed requirements explanations were provided during development process from Product Owner perspective. Software development took some months. It is a good result taking into account integration of OCR modules for paper documents recognition, responsive frontend design for better UX (User eXperience) and integration with ERP software.

Implementation. During software development new user instructions were created with detailed responsibilities and acting scenarios in case of emergency situations. After few test runs on a separate instances of end-to-end processes everybody between personnel was happy to see something new after all these "years of struggle" (c). Full implementation took approximately six months.


IT infrastructure TCO decrease. After sucessful implementation of the new system old ones were rejected. Equivalent of tens thousands of US dollars on annual perspective were saved for the business.

End-to-end business process got more transparency and controllability. Manual labor of "supporting" action for main process dropped almost to zero.

Performance of business increased. Economy of human resources allowed to increase the volume or speed of tasks execution. Hard to give exact numbers, nevertheless about 10-15% of gain was achieved.

Personnel became happier and was ready to accept new business challenges. This was very unpredictable effect. Positive atmosphere on places, transparency and ease of communications helped to reduce stress in company.

TCO of new IT solution was very modest. Around $10 000 per year for 150 users. Approximately 50% of the cost can be treated as "fixed expenses" for IT infrastructure (IaaS model from cloud IT services provider). The rest of amount can be treated as "variable expenses" as they were spent (year by year) on troubleshooting and technical support.


Some images and drafts with hidden details.