CCLA Case Study
CCLA is a medium-scale company based in the City of London, providing specialist investment management to charities, local authorities and faith-based organisations. They invest money for more charities than any other UK fund manager.
As the company experienced significant growth in their client base and found that they need to resolve some issues with their IT infrastructure as a result. The issues they experienced were:
- Disparate systems and data silos storing different versions of client data
- Hard to locate customer documents and information
- Excessive time taken to process instructions
In September 2009, ECS were approached by CCLA and asked to provide a scanning and repository solution. ECS delivered a solution using a combination of ODC (Oracle Document Capture) and Oracle WebCenter Content. This began as a basic implementation, with a thin indexing/classification layer in ODC and WebCenter Content used as an archive for the scanned content.
ODC is a desktop-based scanning/capture and indexing application with some extra capabilities you’d expect, such as barcode recognition, OCR and automatic page separation. The server-client model is very flexible and we initially chose to run machines as independent scanning terminals and servers for maximum availability.
After the successful implementation of the first phase CCLA wished to expand the scope of ODC/ WebCenter Content to provide a richer indexing solution, replacing some responsibilities of existing systems.
ECS deployed their WebCenter Content-based indexing/workflow solution, Iris, which is built around rapid classification and workflow of unstructured content. It presents work in ‘bundles’ of documents, analogous to the contents of an envelope.
Iris is a flexible solution with an open API for integration. After integrating Iris with their existing client records database, it greatly increased their consumption throughput and integrity of captured data, enabling auto-completion and validation of metadata against their static records.
CCLA saw the potential in the ODC/WebCenter Content solution and we developed it further. This led ECS to building WebCenter screens to provide a front-end for their new ‘Central Database’, consolidating client and account information from a set of legacy systems that have since been retired.
ODC has been configured to match their business requirements:
- Automated separation of paper bundles and documents via separator sheets
We built upon ODC’s basic document-level separation, introducing the concept of a ‘bundle separator’ which ensures that sets of related documents are grouped together as bundles following their release to WebCenter Content.
- Barcode recognition on new sets of personalized forms
A combination of WebCenter Content and Create!Form is used to generate pre-populated PDF forms for their client base. Each one features a barcode that is detected and captured by ODC. This provides complete indexing automation in some cases when the forms are returned, along with enhanced tracking of form state.
- Automated upload of client faxes and emails via ODC Import Server
Yielded significant time reduction in processing content from these channels.
The Iris indexing interface has an array of features tailored to fit their needs. It is highly configurable and supports best practices for Financial Services and other industries. The key features are:
- Powerful user interface
A simple interface allows users to access all of the key information on a few key screens. For example a single indexing screen presents the Document with its associated metadata, workflow information and audit log
- Metadata auto-completion
Tight integration between WebCenter Content and the ‘Central Database’ yields powerful features to assist users during indexing of instructions. Key-based document data fields provide a ‘lookup’ button that auto-fills associated fields, providing a match is found.
- Process Date calculation
Transaction instructions must be designated a ‘process date’, indicating when they must be submitted to the Dealing and Registration system to apply movement of assets. Each instruction’s Process Date is calculated automatically, dependent on the dealing dates of the associated funds and the time the instruction was received by CCLA.
- Blind dual-indexing validation to improve data validity
Each instruction type has a configurable set of ‘dual index fields’ – these values are blanked out and must be re-entered by a second user. A comparison check will flag up any bundles where the dual-indexed data fields do not match.
- Prioritization of work bundles
Much of the daily inbound content at CCLA is time-critical and must be processed within days or even hours of receipt. A configurable prioritization system based on content source, transaction data values and date received gives each bundle a score. The score is refreshed after a user updates key field values or finishes processing a bundle.
- Automated assignment of work to users
Rather than manually picking work bundles from a list, users are automatically presented with the next-highest priority bundle to process.
- Bundle flagging and escalation
If a user is unable to process a bundle it can be flagged and escalated to the department administrator.
- Duplicate checking
On completion of bundle processing, a duplicate check takes place and the user is warned about potential duplicate instructions that were indexed previously.
- Integrated signature capture and checking, using fast copy-and-paste upload of new signature images
Certain instruction types allow capture of signature images – typically mandate/registration forms. This process has been streamlined by displaying a list of account signatories alongside the document (with their current image on file, where applicable) with an adjacent ‘upload’ button. This allows a signature to be captured off a PDF directly.
Other instruction types require a signature check to be performed. All captured signature images associated with the current account are displayed alongside the document in Iris, with an option to mark one or more as ‘checked’.
Static Data interface
WebCenter serves as the primary front-end to the Central Database. This stores client, account and person records, along with the relationships between them. Supporting data for these entities is also stored here – contact details, downstream mapping tables, generated form details, campaign enrolments, KYC (Know Your Client) check outcomes etc.
As well as providing the typical create-update-delete functionality against records, the interface has the following extra features:
- Comprehensive search
A single search screen provides easy access to all the core entities. Results are fetched in real-time as you type.
- Relation grids
The core elements are supported by an extensible relations model, enabling different entity types to be linked together in a variety of ways. These are displayed in a grid format, allowing multiple entity relationships to be viewed and updated simultaneously. The relations themselves support their own attributes – these are displayed via colour-coding or by clicking through on a tick icon.
- Experian integration
Person identity checking is seamlessly linked to the Experian Authenticate Pro service – a person record can be checked against a large pool of external sources in a single click, imparting the outcome on screen as a simple red/amber/green ‘traffic light’ icon. Supplementary data, including risk indicators, are available by clicking through on the icon.
Experian address lookup functionality is also embedded in the contact detail views, providing drill-down and auto-completion of address data off a known postcode.
- Static Data Updates
The in-house Dealing and Registration system contains an internal copy of client accounts, bank account details and names/addresses, used for processing creations/liquidations and batch statement generation respectively. Any updates against this ‘static data’ are now controlled exclusively by WebCenter and the Central Database.
On completion of a document bundle, content metadata is converted into an ‘Instruction’ and stored to the database.
An Instruction equates to a specific business task or client directive. Each one is processed via a linked series of modules which “route” the instruction to the desired outcome. Some of these modules perform extra validation checks; others notify downstream systems of updates or tasks.
This extra abstraction layer ensures that instruction processing is not coupled to content items in the WebCenter Content repository. Instructions can be instantiated from other channels without the need for a backing document.
Implementing Oracle middleware has yielded the following benefits for CCLA:
- Virtually eliminated paper instructions held on-site
- WebCenter Content document repository provides easily-accessible, fully-classified history of client documents, correspondence and instructions
- Inbound mail is now processed on the day with no backlog
- Reduced risk of incorrect data capture via auto-completion features
- Data validation/verification and dual indexing features in Iris ensure maximum integrity of captured instructions
- Signature checking and capture integrated with indexing interface, resulting in much faster signature checking
- Automated upload of faxes and emails by ODC eliminates previous manual process
- Bar-coded forms are self-indexing, greatly reducing the time taken to manually process
- Instruction prioritization ensures time-critical instructions are processed first, confirming to client SLAs
- Tight integration with Central Database ensures the majority of captured data for each client/account can be found quickly, and in one place.