BUSINESS SITUATION
Exception management workflow requires human actions for setting up machine learning to enable automated data extraction. Irregular extraction requirements led to high human effort and delay in delivering extracted data to end users. The average cycle time to extract requisite data from a document was approximately 10 minutes, with rapid client onboarding and high ingestion, which made the process not scalable.
SGA Approach
To resourcefully manage the challenges SGA worked with the client product team to develop unified extraction requirements which made it easier to systematically extract data with no human intervention. This helped scale the workflow and helped in the reduction of overall volumes. Furthermore, standardizing extraction strategy with system automation helped reduce average cycle time by 70%
Exceptions Management
Blended Workflow
- 70% automated workflow
- 30% manual workflow
Communication Cadence
- Extraction rate across different document types
- Average processing time per document
- Average TAT for every project accomplished
Process Engineering
- Recommend change in extraction strategies with concerned stakeholders
- Enable machine learning techniques through workflow change
- Workflow automation for reducing cycle time
- Data flow between stakeholders for consistent review and eradication of workflow gaps
Centralized System
- Daily summary reports
- Standardized extraction strategy across projects
- Workflow monitoring tool to enable operational requirements for daily inflow
- Automated summary reports for every project
Automation
- Periodical reports
- Proactive workflows to reduce human intervention and enable machine-learning techniques
- Data Insights for operational outcomes
Engagement
We assessed the AI-based tools and ML capabilities to create a robust, strong exception-handling workflow. This also assisted in enabling proactive measures, reducing manual effort, and replacing it with smart automation.
Benefits & Outcomes
- We achieved 3x growth in volumes with minimum operational costs.
Key Takeaways
- We aided technology with a higher extraction percentage by providing regular data insights.
- We developed a scalable workflow to reduce human intervention and manage multiple projects simultaneously with limited manual effort.
- We implemented proactive workflow management for qualitative operational output.
- We designed diverse operational structures curated for every workflow structure.
- We created a structured cadence to manage agile workflow dynamics.