040ADV-25-R-0008 Questions and Responses to Solicitation - 32025tm.xlsx
XLSX spreadsheet 27 KB Posted
- Attached to
- Detection and Redaction Services
- Solicitation number
- 040ADV-25-R-0008
- Issued by
-
Government Publishing Office
About this file
This file is a Questions and Responses document for Solicitation 040ADV-25-R-0008 for Detection and Redaction Services issued by the Government Publishing Office (GPO). The solicitation seeks a contractor to proactively detect and redact Personally Identifiable Information (PII) within GovInfo Digital Repository collections, primarily processing PDF and digital image files. Key technical requirements include 100% redaction of high-impact PII, maintaining original document structure, supporting multiple languages (primarily English with less than 1% non-English content), and processing between 2TB-15TB of data per performance period, with an initial minimum of 2,556,739 digitally imaged pages.
Pricing is preferred on a per-page basis, with vendors expected to perform discovery of PII in 100% of files and redact only when PII is detected. The contract includes multiple Contract Line Item Numbers (CLINs) covering services like PII discovery, redaction, data transfer, and optional API integration. Performance expectations include using SFTP for file transfer, potential use of removable media for large files, and a 30-business-day window for rework if GPO identifies any remaining PII in deliverables. The solicitation does not indicate an incumbent contractor, and no proposal deadline extension was granted despite vendor requests.
View the file
Other files
| File | Type | Posted |
|---|---|---|
| 040ADV-25-R-0008 Attachment J1- Price Sheet.xlsx | XLSX spreadsheet | |
| Solicitatin 040ADV-25-R-0008-C0001 Amendment.pdf | ||
| Attachment J2 Past Performance Questionnaire.docx | DOCX document | |
| 1) Solicitation - 040ADV-25-R-0008 - Detection and Redaction Services.pdf | ||
| Attachment J1- Price Sheet.pdf |
On GovTribe
Work with this file
- Download the original file
- Contacts named in this file
- Similar government files
- Ask GovTribe AI about this file
Text version
Sheet1
| Responses to Question - Detection and Redaction Services | |||||
| 040ADV-25-R-0008 | |||||
| Number | RFP Reference Section | Question | Response | ||
| 1 | pg 47 Section L1 | Are vendors to submit their proposals to Ms Mizelle and Ms.Webb? | yes | ||
| 2 | Attch J1 | For each CLIN, is the vendor to provide a Min and Max Price Per Year | |||
| (within the "TOTAL PRICE" Columns)? | Yes, please provide Min and Max. An excel spreadsheet will be provided with 2 tabs for min and max. | ||||
| 3 | C.5.1 Item #10 | Pg 37 | For non-English documents, what languages should the system support specifically? | Less than 1% of content is expected to be in Non-English languages. Spanish language content is the most common non-English language to be anticipated. Other languages may include Russian, Chinese, Tagalog, Vietnamese, Arabic, and French. Given the nature of PII and GPO's definition of PII, formal names are the most statistically likely non-English language to be relevant in the purposes of this service. In some situations, there may be known PII and GPO will provide instruction and direction to the known location and context within a document. | |
| 4 | C.5.1 Item #18 | Pgs 37-38 | What are the specific BagIt implementation requirements? | The BagIt specification is provided online https://datatracker.ietf.org/doc/html/draft-kunze-bagit | |
| 5 | C.6.2 | Page 39 | Are there specific security requirements for hosting the redacted content? | Yes, please defer to section C.6.2.2. CONFIDENTIALITY, SECURITY, AND PRIVACY | |
| 6 | C.4.1.2 | Pages 36-37 | What is the preferred review/approval process for redacted documents? | C.4.1.2. Quality Control/Quality Assurance |
Quality Control/Quality Assurance shall be integrated into the service and project implementation. The contractor shall provide a Quality Control/Quality Assurance Plan as part of their proposal. These plans may be written for individual task orders, or a process which accounts, universally, for all task orders in this IDIQ vehicle. GPO has 30 business days from the day of receiving deliverables to provide acceptance or failure of work performed and approval of invoicing. The vendor will perform rework within 30 business days for all deliverables which are not accepted by GPO.
| 7 | - | Will there be a need for manual review of machine-detected PII before redaction? | The vendor is required to implement a quality control methodology to ensure 100% redaction of all high-impact PII. GPO does not prescribe the methodology of how this is achieved. | |
| 8 | - | How should the system handle potential PII that falls into gray areas or has low confidence scores? | All high-impact PII is defined by GPO. PII that does not fall under the scope of GPO's definition of "high-impact" will be considered on a case-by-case basis when discovered by the vendor. | |
| 9 | - | tHow should the system handle corrupted or password-protected files? | The vendor will not receive corrupt or password protected files. | |
| 10 | C.5.1 Item #17 | Page 37 | What are the requirements for maintaining document structure and formatting after redaction? | All original documentation structure (e.g. BagIt) formatting, file features (e.g. PDF bookmars, PDF/A-2b compliance, OCR) shall be retained. |
| 11 | C.5.1 Item #13 | Page 37 | If the file is machine readable, should the output strictly be machine readable? | All redacted deriviates of orginal files will be in the same file format as the original. |
| 12 | C.5.1 Item #4 | Page 37 | Does redaction on images present within documents need to maintain document format? | Yes. |
| 13 | C.5.1 Item #12, #13 | Page 37-38 | In what formats should the redacted data be visible for machine readable files vs scanned/image files? Example should redacted data to be represented as XXXXX or ***** within machine readable files | The specifications for redaction method (e.g. XXXX) will be defined per each task order. Generally, GPO prefers XXXX and does not use "black out" redaction methods. |
| 14 | C.5.1 Item #11, #12 | Page 37 | Are there specific requirements for handling embedded objects within documents? | The vendor is not expected to receive embedded objects within documents. In the rare scenario this occurs, the embedded documents will be seperated and the PDF and its embedded object will be treated as indepdendent files. |
| 15 | B.2 CLIN 0015 | Page 9 | Are there any specific API requirements for potential future integration? | See: Optional 0015-Utilization of GPO’s GovInfo API to “pull” content directly from user interface for PII discovery and redaction. The use of GPO's API, if requested per optional CLIN, will be used solely to 'pull' content from GPO. It will not be used to transfer content to GPO's systems or technologies. |
| 16 | C.4.1.2 | Page 36 | How will GPO validate the redaction accuracy requirement? | Any high-impact PII discovered in deliverables from the vendor will reslult in rework. |
| 17 | C.4.1.2 | Page 36 | What is the expected timeline for rework if issues are identified? | The vendor has 30 business days to provide rework. |
| 18 | C.4.1.2 | Page 36 | Are there specific QA/QC processes GPO expects us to implement? | See C.4.1.2. |
| 19 | B.2 CLIN 0014 | Page 9 | For CLIN 0014, can GPO provide examples of the types of non-PII content that might need redaction? | The content that the vendor is receiving consists of digitally imaged government documents. In some rare situations, these original docment copies contain hand-written information in the margins, or information that might expose personal data of private or public staff. GPO might request that this information is obscured or redacted. It is not expected that the vendor look for, or discover this type of information. It is only expected that the vendor can perform the redaction when GPO makes an explicit request to do so and provides the exact location in the document to perform the action. |
| 20 | B.2 CLIN 0015 | Page 9 | For CLIN 0015, what authentication mechanisms are in place, and are there any rate limits to consider? | Vendors can use freely available keys from https://www.govinfo.gov/api-signup or by reusing an existing api.data.gov key. There are rate limits in place, but they are generous (36k requests/hour) and if needed, adjustments to rate limits can be made on a per-key basis. |
| 21 | C.4.1.2, C.5.1 | Pages 36-38 | What are the performance metrics apart from the 100% PII redaction requirement? | At minimum, the service selected shall perform these activities to scale (across terabytes of content) based on nominal workflows and rule-based systematic approaches to PII removal. Through this IDIQ vehicle, the vendor shall perform redaction of ongoing, simultaneous tasks of varying sizes. Each task may require as little as 2TB or as a great as 15 TB of PDF and other file formats to be processed. Work shall be performed across multiple tasks, and more than one delivery of content may be provided to GPO at a time. |
| 22 | C.5.1 | Page 37 | Are there specific SLAs expected for turnaround time for different volumes of documents? | There are no SLAs that vary from task order to task order, but the vendor is expected to perform work across multple task orders at a given time within the 2TB-15TB of data range over the course of a performance period. |
| 23 | C.3 | Page 35 | Beyond "high-impact PII," are there requirements to detect contextual PII? | There are no requirements to identify nor discover contextal PII. |
| 24 | C.5.1 Item #3 | Page 37 | What is the expected range of task order sizes (in terms of data volume)? | The vendor must accommodate between 2TB-15TB of data per performance period. The first year of performance will, at minimum, require processing 2,556,739 digitally imaged pages. |
| 25 | B.2 | Pages 5-33 | Should pricing be structured per page, per document, or per GB of data? | The vendor is encouraged to represent pricing that provides for the greatest accuracy. Per page is preferred, as PDF formats and their JPEG2000 counterparts will comprise the majority of the work to be performed. It is only expected that JPEG2000 files will be processed if PII is discovered in their PDF equivalent. |
| 26 | B.2 CLIN 0007 | Page 7 | How should we price the data transfer and hosting services (CLIN 0007)? | If there are costs per each task order associated with the hosting of content on local servers and maintainence of the required SFTP site to transfer data, those costs can be reflected in this CLIN. |
| 27 | - | Not explicitly addressed in SOW | Are there any known challenges with similar previous efforts? | No |
| 28 | C.5.1 Item 3 | Page 37 | What are the expected approaches for ingesting files/data and sending the data back to GPO | The vendor is required to implement an SFTP transfer system for GPO to provide and receive files. In some cases, transfer of files will need to be done using removable media (e.g. hard drives), due to large file sizes that can often fail over SFTP. |
| 29 | C.3. Scope, 3. | Page 36 | From the below statement, does it mean that some PDFs were subject to OCR already with low accuracy? |
Page-36
| 3.PDF files provided for redaction may or may not contain searchable text, or they main contain optical character recognized text but not at high accuracy rates. | It is expected that most PDFs will have 99% OCR accuracy, but the vendor is expected to identify PII even the absence of OCR. | ||||
| 30 | Would the government please consider offering a 1-week extension to the current offeror due date? | No, we ill not be extending at this time | |||
| 31 | N/A | N/A | Is there an incumbent on this contract? If so, who? | Contiem | |
| 32 | L. | 51 | The RFP does not include instructions regarding the PPQs. Can the government provide the pertinent instructions? | You are to send the Questionairres to 3 people that you have performed work that is the same or similar as the work requested in this solicitation | |
| 33 | L | 50 | The RFP does not include instructions regarding formatting. Can the government provide guidance regarding font size, font type, graphics, etc.? | single spaced in A4 o r8.5"x11"size paper. Only Times New Roman 11-point or larger fonts must be used. Bolding, underlining and Italics may be used toidentify topic demarcations or points of emphasis. Graphic presentations, including tables, while not subject to the samefont size and spacing requirements, must have spacing and text that is easily readable, and may be multicolored | |
| 34 | L | 50 | Will the government be correcting the numbering issues in Section L? There is no L2. | L.2 VOLUME I – TECHNICAL PROPOSAL (max 30 pages) and L.3 VOLUME II – BUSINESS/PRICE PROPOSAL | |
| 35 | C | 35 | Will the government be correcting the numbering issues in Section C? | No information is missing please disregard numbering issue | |
| 36 | Multiple | Multiple | Could the government clarify the distinction between "offeror" and "vendor" in the RFP? | both are the same individual | |
| 37 | C | 36 | If Contractor has a current ISO 27001 Certification will the Contractor still be required to obtain ISO/TEC 27038 prior to award? | No | |
| 38 | C | 36 | Regarding - "PDF files provided for redaction may or may not contain searchable text, or they main contain optical character recognized text but not at high accuracy rates." Are we to assume that existing OCR and embedded metadata in provided files is not of a quality to be relied upon and thus every page would require OCR | Offerer should assume most PDF content has a 99% OCR accuracy rate but should be able to perform discovery of PII and redaction when OCR is not existing within the document. | |
| 39 | C | 37 | What languages and alphabets other than English may be involved in this project? | Less than 1% of content is expected to be in Non-English languages. Spanish language content is the most common non-English language to be anticipated. Other languages may include Russian, Chinese, Tagalog, Vietnamese, Arabic, and French. Given the nature of PII and GPO's definition of PII, formal names are the most statistically likely non-English language to be relevant in the purposes of this service. In some situations, there may be known PII and GPO will provide instruction and direction to the known location and context within a document. | |
| 40 | C | 37 | What percentage of the documents will be in a language other than English? | Less than 1% of content is expected to be in Non-English languages. Spanish language content is the most common non-English language to be anticipated. Other languages may include Russian, Chinese, Tagalog, Vietnamese, Arabic, and French. Given the nature of PII and GPO's definition of PII, formal names are the most statistically likely non-English language to be relevant in the purposes of this service. In some situations, there may be known PII and GPO will provide instruction and direction to the known location and context within a document. | |
| 41 | N/A | N/A | What is the pace of documents that require redaction on a weekly/monthly/annual basis? | The vendor must accommodate between 2TB-15TB of data per performance period. The first year of performance will, at minimum, require processing 2,556,739 digitally imaged pages. The offerer may provide deliveries on a weekly or monthly basis. The offerer is expected to process ~500,000-1 millions pages per month acorss multiple task orders simultaneously. | |
| 42 | Reference Pricing Table: GPO asks for pricing for Discovery of Text, XML, HTML, or other markup or Unicode standard encoded formats but do not ask for redaction pricing for these formats? | Redaction is included in CLIN 0012 | |||
| 43 | Reference Pricing Table: Why is the quantity of Discovery of PII sometimes greater than the quantity required for redaction? | The offrer will perform discovery of PII in 100% of files received from GPO (unless otherwise directed by GPO to disregard specific files within a directory). The offerer will only perform redaction activities when PII is detected. | |||
| 44 | Reference Pricing Table: There is a note on the bottom of the pricing sheet that says “10% annual increase.” Should we assume all minimum and maximum quantities increase by 10%? | Maximum increases should be expected only. | |||
| 45 | RFP C.5.1, paragraph #5 | 37 | The TO Deliverables requirement lists "other personal information”. Will this data be numeric with predefined patterns like the other PII listed in this paragraph? | This refers to PII that is both numeric or contextual. In the occurance that PII is not defined by GPO's definition of "high-impact," GPO will provide explicit direction on how to perform discovery and, when available, what sections of a document the sensitive information is located. | |
| 46 | RFP C.5.1, paragraph #10 | 37 | Requirement states that non-Latin character codes must be discovered. Please provide a list of languages that the system must be able to discover PII and redact. | Less than 1% of content is expected to be in Non-English languages. Spanish language content is the most common non-English language to be anticipated. Other languages may include Russian, Chinese, Tagalog, Vietnamese, Arabic, and French. Given the nature of PII and GPO's definition of PII, formal names are the most statistically likely non-English language to be relevant in the purposes of this service. In some situations, there may be known PII and GPO will provide instruction and direction to the known location and context within a document. | |
| 47 | RFP C.5.1, paragraph #11 | 37 | Does this requirement mean redacting imbedded images in text-based PDF or Word files? If yes, is this just PII that is identified in the embedded image? Please clarify. | Redaction of images refers to digital images (e.g. JPEG2000 or TIFF formats) which are the master files for which the PDF content is derived from. | |
| 48 | RFP C.5.1, paragraph #12 | 37 | How do we determine if all of the imbedded images are to be redacted or parts. If parts, is just PII that is identified in the embedded image to be redacted? Please clarify. | The offrer will perform discovery of PII in 100% of PDF files received from GPO. The offerer will only perform redaction activities when PII is detected. This inclused redaction of PII in master image file formats. | |
| 49 | Attach J1 | Attachment J-1 is a required submission for Volume II. Does the government have an editable PDF or a MS Excel version they want us to complete? | New Spreadsheet will be provided | ||
| 50 | Attach J1 | Volume II states “The proposal must reflect minimum and maximum cost estimates per each CLIN per each year”. Where in Attachment J-1 only states the Total Price. If the government wants this information in Volume II, the one-page limit is not adequate. If the government wants this information in Attachment J-1, how is this information expressed in the table provided. | Yes, please provide Min and Max. An excel spreadsheet will be provided with 2 tabs for min and max. | ||
| 51 | Section L3. L.3 VOLUME I | The statement “The Management approach shall also include a management plan that discusses, at a minimum, the contractor’s ability to recruit/source, onboard, and retain the talent required to perform on this contract” is in both Factor and Factor 2. Does the government want a duplicate response in both sections or does the government want a single response in Factor 1 or Factor 2? Please clarify. | It is sufficient to provide this information as a single response. Duplicative responses are not necessary nor will they have negative impact on the evaluation of technical factors. | ||
| 52 | Factor 1 | Page 50 | The government states “…The offeror is expected to address each of the enumerated tasks (6 total) in the Statement of Work..” however, please clarify which six tasks as only four tasks are expressly stated in Section C3: 1. Digitally imaged text-based (e.g. PDF) and raster image (e.g. JPEG2000, TIFF) files from historic federal publications which shall be provided to the contractor via an agreed upon file transfer or directory access method . 2. Both digitally imaged and “born digital” PDF files which are currently accessible to the public, but may erroneously contain PII, via the GovInfo digital repository (www.govinfo.gov) which will be provided to the contractor for redaction only after identified as containing PII discovered by the contractor. 3. PDF files provided for redaction may or may not contain searchable text, or they main contain optical character recognized text but not at high accuracy rates. 4. Information which falls beyond the definition of high-impact PII, but is sensitive, or inappropriate for public use, and is identified by GPO, and exact locations within-text are provided by GPO for redaction. | This is a print error, this should reference 18 tasks, as enumerated in C.5.1 TASK ORDER DELIVERABLES |
File details come from the government source that posted it. Updated .