NOVACAP.AI & AIDocBuilder
Novacap.AI is now empowered and improved with AIDocBuilder's Platform. It augments Novadoc's robust offerings and our top-notch IBM content management skill sets with even greater AI functionality to dramatically improve all your automated document processing needs.
De NovaCap oplossing, gebaseerd op het AI Docbuilder platform, is specifiek geconfigureerd om nauw te integreren met IBM Business Automation (Datacap, Filenet, Case Manager), een modulaire set van geïntegreerde softwarecomponenten, ontworpen om werk te automatiseren en bedrijfsgroei te versnellen.
Our Novacap solution
The NOVACAP / AIDocBuilder platform is specifically configured to closely integrate with IBM Business Automation (Datacap, Filenet, Case Manager) which is a modular set of integrated software components, designed to automate work and accelerate business growth.
Input channels are used to digitally capture business-relevant information from various sources in an information management system or an enterprise content management (ECM) system for further processing.
The platform provides flexibility in managing incoming channels from various sources, like email, direct upload, FTP/SFTP, integration with local and cloud storage, and integration with IBM FileNet.
Document preprocessing includes steps taken to clean, structure, and prepare documents for further analysis, modelling, or extraction of information. This essential phase uses natural language processing (NLP), large language models (LLM) and machine learning projects, as it improves the data quality, making it more suitable for algorithms to process.
This module seamlessly integrates with IBM Datacap through web services, which can provide robust imaging tools and multiple recognition engines.
Document classification assigns predefined categories or labels to a document based on its content. It is a common AI task that organizes, filters, or sorts documents automatically. The platform includes pre-trained classification models for standard document classes, like Invoices, Packing lists and transportation documents.
The Data Extraction module uses artificial intelligence techniques to automatically extract information from documents, emails, images, or web pages. Our AI data extraction module leverages modern machine learning techniques to identify and pull-out specific pieces of information, such as key-value pairs, address blocks, and named entity recognition.
The module includes very powerful detection of tabular data, identifying rows, columns, and the data they contain for very complex tables inside commercial documents.
This process significantly reduces manual effort and improves accuracy, speed, and efficiency in handling large volumes of documents.
Data assembly automatically gathers, organizes, and structures data from multiple sources into a coherent dataset for analysis, machine learning, or decision-making. This process aims to collect disparate, unstructured, or semi-structured data, harmonize it, and prepare it in a format that downstream systems, such as business intelligence tools, document management databases or any other data repositories, can efficiently utilize.
The module can gather raw data from multiple sources, including databases, documents, web APIs, and other online or offline data repositories and combine data into a unified dataset formatted as JSON, XML or CSV.
API Manager module is built as a stand-alone dataset or model export service. The module oversees the entire lifecycle of APIs running through the platform. It can interact directly with cloud storage like AWS or Azure, allowing the export of datasets and documents, whether preprocessed, cleaned or feature engineered. Pre-built connections can export data and documents to IBM FileNet, IBM Case software, and any database.
If the export operation is time-consuming, the API supports asynchronous job creation with mechanisms to check the job’s status. It informs the user when the export job is completed or failed.
QUESTIONS? READ OUR FAQ!
Use Case for Trade Finance
As documentation organizing plays a vital role in international trade, missing or incorrectly prepared documents pose a risk for both buyers and sellers, as this can cause delays in shipments and, ultimately, delays in payments.
Trade Finance transactions involve extensive documents, ranging from invoices and bills of lading to certificates of origin. The complexity of documentation can lead to delays and errors, causing disruptions in the supply chain. Using our platform can streamline the documentation process, reduce the risk of errors, and expedite the overall trade cycle.
Documentation complexity in trade finance arises due to the involvement of multiple parties, regulatory requirements, and the nature of the transactions, which often cross international borders.
The transactions typically involve stakeholders, including exporters, importers, banks, insurers, freight companies, and customs authorities. Each party requires specific documentation to confirm their part in the transaction.
Document Types Processed in Trade Finance
-
Commercial Documents: Contracts, purchase orders, commercial invoices.
-
Transport Documents: Bills of lading, airway bills, packing lists.
-
Financial Documents: Letters of credit, guarantees, promissory notes.
-
Regulatory Documents: Customs declarations, certificates of origin, compliance reports.
Different countries have specific regulatory requirements and compliance obligations for trade finance transactions. Ensuring that all documentation meets the legal standards of multiple jurisdictions adds to the complexity. Some examples include:
-
International laws and standards (e.g., ICC rules like UCP 600 for letters of credit).
-
Local customs laws and regulations.
-
Anti-money laundering (AML) and counter-terrorism financing (CTF) requirements.
This industry aims to mitigate the risk of non-payment or non-delivery of goods. Complex documentation, such as letters of credit and performance bonds, are required to ensure financial and delivery security. The specificity of these documents must align with each party’s expectations and the transaction terms.
Many of these processes are still paper-based, which adds to the difficulty of managing and verifying large volumes of documents. This often results in delays, human error, and additional costs.
Most of the time, the documentation is highly time-sensitive. Delays in submitting or processing key documents, like a letter of credit, can delay shipments or even cause a transaction to fail. Ensuring the timeliness of all required documentation across different stakeholders can be a significant challenge.
Different institutions may interpret documentary requirements differently. For example, one bank might accept a document format that another rejects. This leads to additional time spent reworking and resubmitting documents to meet varied standards.
The complexity of documentation in trade finance also opens the door to potential fraud, such as falsifying documents or misrepresenting goods. Banks and institutions must have robust verification and due diligence processes, adding further complexity.
References
A large shipping enterprise was challenged by poor customer support and low customer satisfaction scores around the processing of their customers’ transportation and financial orders documents - they simply didn't have the tools to manage these custom workflows for their customers. They had an existing IBM Business Automation environment and were struggling to automate these flows.
They were also plagued when onboarding new customers - their needs were so complex that it took 4 full-time employees 4-6 weeks each time to onboard 1 customer. And even then, the functionality was very minimal; they could only do basic indexing of invoices - they only had some key values and could only process a simple table.
The best-case scenario was that, once a new customer was onboarded (which itself took over a month), it took 5 to 10+ mins to fully process an order, including uploading documents. Even when they started processing orders, it involved multiple people to monitor it and address any mistakes or changes.
Solution - after
Fast forward to today, with NOVACAP, they are utilizing a platform that onboards their customers in less than 5 minutes and, once onboarded, invoice processing time went down from 5-10+ minutes to 3 seconds; all done automatically, and without IT's assistance. A business user is trained in 1-2 hours and then is self-sufficient. And they reduced the onboarding team from 4 to 1!
They also improved their processing volume to the point where they now handle over 100K financial and transportation documents processing automatically every month. In contrast, before, they could only manage a few thousand and with abysmal customer satisfaction.
They even looked at several other content extraction and automation platforms but those could not handle the complexity or demanding volume requirements. Also, the other products' automation capabilities were weak and did not address many different document types they needed. They were especially poor at handling complex tabular documents which is a standard format to deal with in this industry.
Another problem they had was that these custom scenarios involved IT every time, and even with their assistance, the 13+ process steps were primarily manual as there was only 60 – 80 % of document data extraction capability.
Lastly, before working with NOVACAP, they could only address a handful of countries that they needed to support due to the tool's poor language coverage.
This process went from 13 (many manual) steps to 4 that are now automatic, with 100% content extraction of their complex tabular documents. As a result, their customer satisfaction scores are now fantastic, and they are now supporting their customers’ processing needs in over 70 countries and counting.
Finally, regarding labour costs alone, they estimate the NOVACAP platform saves them over $1 million annually!
Large shipping enterprise
Challenges - before
Forbo Eurocol Nederland produces various types of floor coverings and related products such as adhesives, levelling compounds, maintenance and cleaning agents. The company receives around a hundred invoices per week, both by post and digitally. Previously, these were processed manually, which regularly resulted in invoices being lost or left on employees' desks (for too long).
Fast and simple
Initially, Forbo Eurocol had hired another partner. After a year and a half, they decided to part ways with that partner. De Ridder: 'Their program had too long a processing time, crashed regularly and was therefore very unreliable. In 2014, we started working with Novadoc. The system is so simple that when there is a problem, it is usually something I can solve myself quite easily.'
Since Forbo Eurocol has been working with Novadoc's NovaCap capture solution, the storage of incoming documents has become much more efficient and faster, and significantly fewer errors are made. Martijn de Ridder, Assistant Controller at Forbo Eurocol: 'Where we used to work with a cabinet full of folders, all invoices are now stored digitally and are therefore much easier to find.
De Ridder: 'My colleague who processes all the invoices used to spend at least an hour a day on this. Now it's half an hour at most. But I think the biggest gain is the fact that everything is neatly stored digitally and can be retrieved very easily. If we now get a question about an invoice, we can answer it immediately, without first having to walk to a cupboard in another department.'