Revolutionising Invoice Processing: Airtel Digital’s Advanced Document Digitisation AI Tool
In today’s fast-paced business world, organisations are constantly seeking innovative solutions to streamline their operations and enhance efficiency. One such breakthrough at Airtel Digital is our cutting-edge document digitisation capability designed to tackle the manual validation challenges associated with processing invoices. By harnessing the power of Document AI and integrating Robotic Process Automation (RPA), we’ve transformed the invoice validation process, reducing manual efforts and errors significantly.
Challenge Overview:
In the Finance P2P (Partner portal invoice validation) workflow, the reconciliation of vendor bills is currently a manual process. This step is crucial for the business, given that approximately 1 Lakh+ invoices (40k Portal based invoices, 60k Non-Portal based invoices) undergo validation each month. The sheer volume of invoices necessitates a considerable manual workforce, making the process error-prone and posing potential audit issues.
The Integration of RPA and Document AI:
At Airtel Digital, we understand the importance of automation in handling repetitive tasks. Our Document digitisation tool seamlessly integrates with RPA, where invoices are first extracted from various systems by the bots. RPA acts as the first line of defence, capturing data from disparate sources and preparing it for further processing. This not only accelerates the overall process but also ensures that the input data is standardized and ready for OCR analysis.
Document AI Role in Processing Invoices:
Once the invoices are handed over to Document AI capability, the real magic begins. The OCR tool meticulously analyses invoices in various formats, including PDF and JPEG, extracting essential information with a high degree of accuracy. This not only eliminates the need for manual validation but also significantly reduces the chances of errors associated with human intervention.
Our central data science team has achieved a remarkable milestone in developing an in-house capability that stands as a testament to innovation and forward-thinking. This is not just any tool; it’s a multi-modal AI model that is intelligently designed to understand the nuances of document text positioning, word semantics, and types. What sets it apart is its adept use of OCR (Optical Character Recognition) combined with NER (Named Entity Recognition), seamlessly working together to process information with unparalleled precision.
One of the most notable feats of this tool is its ability to distinguish between scanned images and true PDFs — a task that often poses significant challenges even for some of the most advanced paid tools in the market. This discernment is crucial, as it allows for tailored actions to be taken for each type of document, thereby enhancing the accuracy of the results obtained.
But there’s more — our tool smartly categorises documents based on their specific types and document type variants. It recognises that an energy bill is inherently different from a credit note and processes each accordingly, demonstrating a level of sophistication in document handling.
At its core, this capability is trained to meticulously identify and extract key data points such as invoice numbers, dates, line items, and totals. This ensures that every piece of vital information is captured, leading to a comprehensive understanding and processing of each document. It’s not just about reading text; it’s about understanding the essence of each document in its entirety.
User-Friendly Dashboard for Analytical Insights:
Airtel Digital recognises the importance of providing users with intuitive and insightful tools to monitor and manage the OCR and RPA processes. To achieve this, we have developed a user-friendly dashboard accessible through a UI portal. This dashboard serves as a centralized hub for tracking the progress of the OCR tool and RPA processes in real-time.
Key features of the dashboard include:
Process Analytics: Visual representation of the entire invoice processing workflow, allowing users to monitor each step and identify potential bottlenecks.
Error Tracking: Detailed insights into any discrepancies or anomalies detected during the OCR and RPA processes, enabling quick resolution and continuous improvement.
Performance Metrics: Key performance indicators (KPIs) such as processing time, accuracy rates, and throughput are displayed, providing a comprehensive overview of the system’s efficiency.
Customisable Reporting: Users can generate customized reports to meet specific business requirements, facilitating data-driven decision-making.
UI Portal as the Conduit:
The UI portal serves as the conduit for seamless communication between the Digitiasation tool, RPA processes, and end-users. APIs provided within the UI portal enable easy integration with external systems, ensuring a smooth and cohesive workflow. This user-centric approach enhances the overall user experience and encourages widespread adoption within the organization.
Overall Solution Review:
RPA Invokes: RPA takes the lead by initiating Digitisation step post-invoice retrieval from the Partner portal. This automation eliminates the need for manual intervention, thereby streamlining the process and promoting automation.
Document AI Evaluates: Document Digitisation AI not only scans and categorizes but also extracts information and feeds it back to RPA, enhancing overall efficiency.
Speed Scanning: With an impressive average processing time of around 6 seconds per bill, this solution significantly accelerates the validation process, addressing the need for speed.
Analytical UI: The system’s modular design and user interface capabilities not only enhance the current workflow but also pave the way for the seamless onboarding of new projects or teams in the future. This approach contributes to the creation of a Unified App, ensuring a cohesive and scalable solution for diverse needs.
Conclusion:
Airtel Digital’s Document digitization capability, integrated with RPA and presented through an intuitive UI portal, represents a significant leap forward in addressing the challenges associated with manual validation of invoices. By automating tedious and error-prone tasks, our solution not only improves operational efficiency but also empowers organizations to make informed decisions based on accurate and timely data. As businesses continue to evolve, embracing innovative technologies around Document AI and RPA becomes imperative for staying ahead in the competitive landscape.
Deep appreciation for the exceptional individuals who transformed this vision into reality. The relentless dedication and steadfast commitment of the team — Devi Khatua, Shruti Jain, Krishan Chawla, Sandeep Grover, Vijay Sharma, Aditya Jain, Rajdeep Chauhan — under the insightful guidance of our esteemed leaders Soumen Ray, Pradipt Kapoor, Ravinder Bansal, Abhishek Singh, and Sudha Gupta have played a crucial role in bringing this groundbreaking initiative to fruition.
