Saztec takes great care to design data conversion methodologies that will result in a high quality product, produced quickly, at the lowest possible cost. One of our project managers will work with you to determine the requirements of your data conversion project. As a result of this, a project plan and detailed technical conversion specifications will be written that will document the methodology chosen for your particular project. Below are some examples of the methodologies that Saztec uses.
Back file Conversion
The generic workflow for a legacy document conversion project includes the following parts:
| Project Setup | A member of the project management team will work with you during the analysis and information gathering portion of the project. The project manager will document the resulting information and conversion requirements in a technical data conversion specification. The software development team will then write any programs necessary to fulfill your data conversion requirements. |
| Testing | After the workflow process has been created and the necessary programs are written, the production team will be assembled and trained in the detailed project requirements. You will be asked to supply a set of representative test documents and these documents will be converted. The resulting data will be delivered for your inspection and acceptance. Meanwhile, Saztec's production management team will evaluate the feedback obtained on the workflow process during testing and make necessary adjustments. |
| Document Preparation | After you and the Saztec production management team declare the test to be a success, the production work will begin. The first step in the production process is document preparation. This includes all of the work necessary to prepare the source documents for further conversion, such as removal of staples, paperclips and binding, the creation of batches, and the insertion of batch sheets or bar code sheets. |
| Scanning | This involves the scanning of source documents in order to create raster image files. Options for the creation of images include; resolution (DPI), format, and depth (bits per pixel - black & white, grayscale, color). |
| OCR | Optical Character Recognition (OCR) is used to create text data. Data can be output in a number of text or word processing formats. The text data that is output through OCR can be verified - where the text is manually reviewed and character errors are corrected. |
| Data Entry | Saztec's "key, key, compare" process provides character accuracy rates of up to 99.995%. Data entry is necessary when text data is required, but the source documents are not good candidates for OCR. Text data can be keyed and output into any format necessary including; XML, SGML, comma delimited, full text, etc. |
| Quality Control | Quality assurance is a necessary design aspect of the conversion process workflow. Throughout the conversion process, data integrity and quality is checked. We also conduct quality checks on final output data to insure that project quality objectives are met prior to delivery of data. |
| Packaging | Data is packaged into the format necessary to be loaded into your computer system. This includes directory naming, file naming, file format, and formatting of output media. |
| Delivery | Data can be transmitted via the internet, including; FTP, email, message queue, etc., or it can be written to output media and shipped. |
Litigation Support
Our Litigation Support workflow includes the following steps:
| Receiving | Paper documents collected during discovery are boxed and shipped to Saztec. Electronic documents can be transmitted to Saztec, or shipped on CD, diskette or magnetic tape. |
| Logging | Boxes, documents and electronic files received by Saztec are logged into Saztec's tracking database. |
| Document Preparation | Documents are removed from the box, bates stamped and prepared for scanning. This includes removal of staples, paper clips and other binding material. |
| Scanning | Each page is scanned in order to create an image file. Options for the creation of images include; resolution (DPI), format, and depth (bits per pixel - black & white, grayscale, color). |
| OCR | Optical Character Recognition (OCR) is used to create text data. Data can be output in a number of text or word processing formats. The text data that is output through OCR can be verified - where the text is manually reviewed and character errors are corrected. |
| Electronic File Conversion | Electronic files, including email files, are converted to the required format. |
| Data Entry | Saztec's "key, key, compare" process provides character accuracy rates of up to 99.995%. Text data can be keyed and output into any format necessary including; XML, SGML, comma delimited, full text, etc. |
| Quality Control | Quality assurance is a necessary design aspect of the conversion process workflow. Throughout the conversion process, data integrity and quality is checked. We also conduct quality checks on final output data to insure that project quality objectives are met prior to delivery of data. |
| Packaging | Data is packaged into the format needed by your litigation support, case management or trial presentation software. This includes directory naming, file naming, file format, and formatting of output media. |
| Delivery | Data can be transmitted via the internet, including; FTP, email, message queue, etc., or it can be written to output media and shipped. |
Resume Processing
The generic workflow for a resume conversion project includes the following parts:
| Receiving | Email: One or more
e-mail addresses can be created on SAZTEC's web server, and
Saztec will check each email "box" daily for incoming
resumes. Each email address can be associated with specified
source codes.
Fax: SAZTEC maintains equipment and telephone lines to accommodate the receipt of facsimile resumes. Each fax number can be associated with specified source codes if requested. PO Boxes: SAZTEC can open one or more post office boxes as required for the receipt of resumes, and will pick up mail from the post office daily for processing. Couriers: Resumes can be shipped directly via the US Mail, or courier to SAZTEC. We supply a Batch Cover Sheet that allows for the indication of source codes and job req numbers. Source Coding: Source codes and job req numbers can be specified at either the batch or the resume level. Business rules should be established to control the placement of coding information on emails, faxes and hard copy resumes. Saztec can insure the use of valid values in these fields if given a validation list for each field. |
| Document Preparation: | Hard copy documents are prepared for scanning by removing paper resumes from envelopes, and removing any staples, paper clips, etc. Email resumes are viewed and any "junk" or "failing" resumes can be rejected and returned to a specified email address. All resume batches are logged into our document tracking system. |
| Image Files: | Paper resumes are scanned to create a 300 DPI (dots per inch) images in TIFF (tagged image file) format, compressed using Group 4 compression. Images can be converted to different formats as requested, such as PDF, JPEG, etc. Output files can be named as required. |
| Index File Creation: | One index file can be created
for each resume in a format as required, for example, XML,
comma delimited, etc. Common fields to be captured include;
name, address, phone number, email address, source code, job
req number. Output files can be named as required.
Two different operators will key each resume. The two input files created independently are compared to locate and correct mismatches. This process works on the logic that the probability of two independent working persons committing the same error is almost zero. Validation and edit checks are then run using validation lists supplied by the client. Errors are corrected, and the final files will have a character accuracy rate of 99.95%. |
| OCR Full Text File Creation: | For email resumes, each mail message and the corresponding attachments are converted to one text file. For paper resumes, full text files are created by processing each image with OCR (optical character recognition) 5 engine, voting software that produces extremely high quality output. Manual cleanup ("Verification") can be applied to resulting text files if requested. |
| Packaging and Data Delivery: | Output files and directories
can be structured and named as requested. Data can be
delivered via transmission, via the a password protected
directory on Saztec's ftp site, or on output media, such as
CD and shipped via courier.
A printed response card or letter can be sent for each resume processed. Email messages can be sent to acknowledge email resumes processed, or to notify applicants if their resume submitted via email was "unprocessable". Data from the tracking database can be output to reports as requested, including; Daily Status Reports, Invoice Detail Reports, etc. Hard copy resumes can be returned to a specified address, stored for 3 months or less, shred, or disposed of. Backups of electronic data are retained for 3 months. |