Manual and Automated Data Entry
If you’re searching for reliable and accurate data entry and verification services combined with security and confidentiality guarantees, then CPI Data Services can be your partner. Your project will be handled by our experienced process management staff to deliver high accuracy output (up to 99.95%) using a variety of quality techniques:
Our Data Entry Process
Every document indexing and coding project is unique. The CPI project management staff works with you to identify the indexing and coding field requirements and the optimal, most cost efficient, high quality process possible.
Pre-planning
Each project begins with thorough, detailed project analysis and testing.
At the onset, CPI’s U.S.-based staff meets with each customer to address project goals, desired outcomes, possible barriers to success and integration with internal processes.
Following this discussion, the team will map the cost efficient process, establish service metrics, identify system requirements, identify monitoring and reporting requirements, and determine resource needs (skills, experience, and training).
Decision: Automated or Manual?
Indexes can be automatically created by applying standard technology such as bar codes or OCR to extract indexing data for standard forms. For images with poor quality, illegible characters, unstructured documents, or those needing interpretation, a custom developed capture process may be necessary.
CPI will recommend a best process methodology for the project requirements:
Single/Double Entry Keying:
Another voting method where two or more operators type the same data. Electronic comparison spots any differences for correction or client notification of illegible documents.
OCR/ICR Voting Technology:
Two or more OCR engines must agree on character validity or flag the character for manual verification
Verification Software:
Using business rules, field validation, table lookups, consistency formulas and logic rules. Validation tables assure that only acceptable data is entered and spelled correctly. For example programs check Zip Code, city and state against Postal Service database or match medical diagnostic codes against International Classification of Diseases database. Custom validation rules and checks can be set.
Logic Rules:
This is automated common sense, in effect. A Social Security number must have nine numbers, for instance. Data must accurately sum or calculate in different sections of forms or reports. Dates must meet logical ranges and sequences. Anything else is flagged for checking. Spell-checking and proofreading software, as well as humans, are used when appropriate.
Random Checks:
We also perform spot checks of data entry accuracy, proofreading and visual inspection of output documents or databases, and data sorts with logic rules. In addition, each staff member’s accuracy is monitored to improve training and reduce errors.
Next Steps
An Agreement with project terms is prepared. It includes a warranty period, confidentiality pledge, service level expectations (measurements, reports), control factors and points of contact. Also addressed are emergency or problem escalation contacts and process, as well as responsibilities, risk-sharing, errors and omissions insurance.
Naturally, project agreements also document compensation, incentives and penalties.
Beginning and Completing the Project
After staff training and testing, parallel processes are run until customer QC approves CPI outputs. Then, production begins. Analysis and reporting continues during project, with adjustments when needed.