Setting up autoconfiguration for a complex Fintech client

Our global financial services provider was experiencing time issues when onboarding new clients. It took up to 11 months to onboard a new client into their services platform which created a backlog of clients for implementation. They wanted us to reduce customer onboarding time by 7 months.

We approached this issue with a two-step solution.

Configuration automation: First we created as-is configuration automation. Then we extracted the client’s configuration worksheet requirements and put them in a structured format such as JSON. Next, we converted the client’s configuration files into XML and DB-script. We then used these files to set up the financial institution instance.

• Developed a process for autoconfiguration using the client’s XML and DB-script configuration files.
• Extracted the requirements from the configuration worksheets and put them in a structured format such as JSON.
• Developed a rules engine to convert the structured requirements into the client’s proprietary XML files.
• Developed the basic financial institution instance setup details into DB SQL scripts.
• Loaded the client’s XML files and SQL scripts into the configuration portal.
• Verified the financial institution client instance.

Once, completed, we moved to the next step in the process.

Requirements automation: We set up a web portal to gather the client requirements. We then converted them into the client XML and DB-script client configuration files. We used those files to set up the financial institution instance.

Here are the steps we took:

• Developed a requirement portal to collect quality requirements and enable autoconfiguration.
• Designed data templates that support the client’s complex business logic for both nested and dependent requirements.
• Developed a responsive web portal that is driven by data templates.
• Built and added in-line error validations to check the data quality entry.
• Integrated with the rules engines to enable autoconfiguration.

As a result of this meticulous process, we identified the areas of the automation process that could help reduce the onboarding timeline. In the process, we brought innovation to our client, and they now enjoy the latest technology and enablers. Data that is entered into the requirement portal is automatically validated based on the client’s business rules and is configured behind the scenes. To improve quality configurations and reduce the defects, the rules only allow users to save valid data in the system.

Our financial services client had more than 50,000 product contract documents that we needed to extract data from. The data had to be extracted in a usable format and then migrated to their contract management system. Additionally, the client wanted us to develop dashboards on the extracted data for ongoing maintenance.

We began the process by sorting the files. The client teams shared all of the contract documents through secured FTP. Then, the offshore teams downloaded the documents on a US server. The cognitive extraction team read the documents and converted them from image to TEXT formats. Once the documents were sorted, we grouped them into client folders.

Next, we worked on extraction. For read easy entities, we used cognitive extraction. We created a document extractor (web application) to render the documents for manual users. And the complex entities were manually extracted. We next performed a quality review. Our QA team performed an audit check for correctness and completeness. We also performed a spot check on the final reviewed documents to uncover any errors. If they existed, we corrected them and saved the updates. We also created dashboards and PowerBI reports on the data extracted.

We repeated this process for each of the 50,000 contract documents. Once we had uploaded the final extracted documents, we asked our client for a review and signoff.

In all, we extracted 22 data entities from 50,000 contract documents that had 98 distinct document types—and we did it within 10 weeks. We achieved 100% accuracy and re-grouping. Also, we developed standard PowerBI reports for the client’s ongoing maintenance.