AIAPT RFI 020425 FINAL.docx

Department of Veterans Affairs (VA) Office of Procurement, Acquisition, and Logistics (OPAL) Strategic Acquisition Center (SAC) Acquisition Integration Service (AIS)

Request for Information (RFI) Artificial Intelligence Assisted Procurement Tools (AIAPT)

February 4, 2025 Version: 1.0

VA OPAL SAC AIS

Request for Information (RFI)

NOTICE

The Department of Veterans Affairs (VA) is issuing this RFI in accordance with FAR 15.201(e). The agency does not intend to award a contract but rather gather information on capability, delivery and other market information pertinent for acquisition planning. There is no solicitation at this time, the result of this market research may contribute to determining the method of procurement, development of requirements and identify parties having an interest in and the resources to support this requirement.

Table of Contents

VA OPAL SAC AIS

Request for Information (RFI)

Page i

PART 1: SOURCES SOUGHT BACKGROUND AND OBJECTIVES

Purpose The Department of Veteran Affairs (VA) Office of Acquisition, Logistics, and Construction (OALC), Office of Procurement, Acquisition, and Logistics (OPAL), Strategic Acquisition Center (SAC), Acquisition Integration Services (AIS) is conducting market research only to determine the availability of qualified vendors capable of providing a proven commercial-off-the-shelf (COTS) generative artificial intelligence (GenAI) solution capable of delivering state-of-the-art large language model (LLM) capabilities in support of acquisition and contract lifecycle management functions. The solution must provide a commercially proven base platform with an integrated suite of robust analytical tools and GenAI capabilities that may be customized to meet specific SAC requirements and use cases. The solution must support a range of AI-enhanced workflows to assist in acquisition planning, document generation, pre-award and solicitation, evaluation and award, and post-award management workflows. The solution must meet all current Trustworthy AI requirements, meet federal security requirements found in VA 6500 Handbook (https://www.va.gov/vapubs/viewPublication.asp?Pub_ID=1255&FType=2), and have a current moderate level or higher FedRamp-authorization for commercial software solution. Additional specific capability requirements are enumerated below.

SAC AIS is not, at this time, seeking proposals and will not accept unsolicited proposals. All responses shall be solely utilized for information and planning purposes. All information received in response to this RFI that is marked “Proprietary” will be handled accordingly. Responses to the RFI will not be returned. In accordance with FAR 15.201(e), responses to this RFI are not offers and cannot be accepted by the Government to form a binding contract.

Responses to this RFI will assist the VA in determining the potential level of interest, competition adequacy, and technical capabilities of commercial vendors to provide the required products and/or services. The Government does not guarantee any action beyond this RFI.

The Government is under no obligation to issue a solicitation or to award any contract based on this RFI. All costs associated with responding to this RFI will be the sole responsibility of the respondent.

Background SAC AIS strives to provide high quality, effective, and efficient integrated acquisition, and business solutions necessary to fulfill VA’s mission.

AIS is seeking input from industry regarding their capability to provide market- leading solutions capable of meeting the full breadth of functional requirements outlined in this RFI. SAC requires access to a platform that is compliant with current Federal security requirements, is easy to use, flexible, scalable, customizable, and configurable to meet specific VA and SAC requirements as further detailed below.

RFI - Information Sought As further detailed in Section 2, SAC AIS is seeking to identify capable and interested partners able to provide the products and/or services required to meet the full range of functional capabilities outlined in this RFI. Viable partners shall have demonstrated experience deploying the solutions and delivering the ongoing support required to maintain, expand, and upgrade the solution over time for an organization of similar size, scope, and mission complexity as VA.

Information requested in this RFI includes the following categories:

Company Information (Section 1) Identification of interested vendors, socio-economic status, and potential acquisition vehicles to inform VA’s potential acquisition strategy options.

Capability Assessment Against Functional Requirements (Section 2) Description of vendor’s commercial products and ability to fully meet all stated requirements. Provide a response for each of the key categories listed below.

I. AI System Capabilities and Integration This category evaluates the specific AI technologies a vendor's system uses, such as machine learning or natural language processing, and assesses how well these technologies fit the VA's acquisition needs. It also examines the system's ability to integrate seamlessly with existing VA infrastructure and handle scalability as the volume of contracts grows.

II. Data Privacy, Security, and Compliance This category focuses on the vendor's ability to safeguard sensitive procurement and vendor data through robust security measures like encryption and access controls. It also assesses the system's compliance with federal and VA-specific regulations on data protection, cybersecurity, and procurement ethics.

III. Bias, Fairness, and Transparency This category examines the vendor's strategies for preventing bias in AI models, ensuring that the system promotes fairness in vendor selection and decision-making. It also evaluates the transparency of AI-driven decisions, allowing VA staff to understand and oversee these processes, while adhering to ethical standards.

IV. Accuracy, Reliability, and Human Oversight This category evaluates the accuracy and dependability of the AI models, particularly for critical tasks like vendor selection and contract management. It also examines the mechanisms in place for human oversight to intervene in AI-driven processes, especially in high-value or complex contracts, as well as the frequency of model updates and validation.

V. Cost, Implementation, and Vendor Support This category considers the short-term and long-term costs associated with implementing the AI system, including training, maintenance, and updates. It also assesses the level of support the vendor provides for system integration and ongoing maintenance, and the measures taken to prevent vendor lock-in to ensure flexibility for future integrations. For example:

VI. Risk Mitigation and Contingency Planning This category reviews the vendor's strategies for mitigating risks such as data integration challenges and handling potential data silos or system incompatibilities. It also evaluates the vendor’s contingency plans for AI system failures or technical issues that could disrupt procurement operations, and the system's adaptability to evolving legal and regulatory standards.

Past Experience (Section 3) Past project example(s) demonstrating vendor’s experience implementing their products to address the needs of an organization of similar size, scope, and mission complexity to VA.

Submission (Section 4) Instructions for submission format, page limits, and required document specifications to ensure consistency and compliance with VA’s response requirements.

Page iii

PART 2: SOURCES SOUGHT REQUESTED INFORMATION

Respondents are asked to provide the information requested below the instructions outlined in Section 1.3 above. Submissions not adhering to the submission instructions may be deemed non-responsive and may receive no further consideration.

Section 1 - Company Information Please provide the following details regarding your organization:

1. Company name

2. Company point of contact (POC) name.

3. Company POC phone number

4. Company POC email address

5. Company URL/web address

6. Company Cage Code

7. Company UEI Number

8. Company business size in terms of revenue and number of employees

9. Company Federal socioeconomic classification(s), if any, based on North American Industry Classification (NAICS) Code that applies to the solution.

10. List GSA schedule, Government wide acquisition contracts (GWACs), or other Federal IDIQs or contracts the Respondent holds that are relevant based on size, scope, and complexity as compared to this requirement and accessible by VA, including corresponding schedule number or contract number.

Section 2 – Functional Requirements Assessment VA SAC’s functional requirements for the solution are outlined below. Respondents are asked to provide a structured and annotated response that clearly maps the description of the proposed solution’s capabilities to the numbered requirements below. The Respondent’s narrative should provide sufficient details regarding the solution’s technical capabilities to demonstrate that the solution is capable of fully satisfying each stated requirement. Responses that do not clearly map to the information stated in the listed requirements below, or are otherwise unclear, may be deemed non-responsive and may not receive further consideration.

Mandatory Qualifications The following qualifications and capabilities are mandatory requirements for any potential solution addressing the needs outlined in this notice. Respondents should assess these “Go/No-Go” requirements carefully, as solutions unable to meet these requirements will not be considered as viable options and will not receive further consideration:

a. Solution must be based on a commercially proven COTS software platform – VA SAC is not seeking custom-developed solutions that have not yet been successfully deployed at other Federal agencies similar in size and scope to VA.

b. Solution must possess significant out-of-the-box functionality enabling it to be deployable on day one of the project, while also supporting custom configuration within expedited timelines.

c. Solution must be built as a single, vertically integrated platform with end-to-end solutions leveraging seamless integration across capabilities and tool sets.

d. Solution must incorporate comprehensive in-house professional and configuration services, supported by a cohesive and specialized team of experts.

e. Solution must comply with all VA directives developed in accordance with FISMA, HIPAA, NIST, and related VA security and privacy control requirements for Federal information systems. Any software solution must be VA Technical Reference Model (TRM) compliant. More information can be found here: VA Technical Reference Model

f. Solution must have moderate level or higher FedRAMP-authorization for a commercial software solution.

g. Solution must be capable of encrypting data, both at rest and in transit, via mechanisms compliant with all applicable Federal and VA Information Processing Standards

h. Solution must be able to comply with all applicable Federal and VA Cybersecurity Supply Chain policies.

AI System Capabilities and Integration This category evaluates the specific AI technologies a vendor's system uses, such as machine learning or natural language processing, and assesses how well these technologies fit the SAC's acquisition needs. It also examines the system's ability to integrate seamlessly with existing VA infrastructure and handle scalability as the volume of contracts grows.

1. Specific AI Technologies: The system must employ specific AI technologies (e.g., machine learning, NLP) and apply them effectively to acquisitions.

2. Integration with SAC Systems: The system must integrate with SAC acquisition systems, including standards such as HL7 and DICOM.

3. Scalability: The system must handle increasing contract volumes without performance degradation.

4. Collaboration Tools Integration: The system must integrate with MS Office Suite and Teams to support real-time collaboration, version tracking, and workflows.

5. Dynamic Document Generation: The system must automate the creation of SOW, SOO, and PWS documents using AI templates.

6. Procurement Data Integration: The system must integrate procurement data into solicitation documents.

7. AI Training Modules: The system must include AI-powered training modules or NLP tutorials to assist requesters in completing documentation.

8. Automated Requirements Collection: The system must use AI (e.g., OCR, NLP) to collect, extract, and validate requirements documentation.

9. Accuracy and Completeness: The system must ensure the accuracy and completeness of purchase request documentation through compliance checks.

10. Dynamic Document Leverage: The system must leverage SOW, SOO, or PWS documents to support risk management, performance tracking, and other administrative functions, with examples of cross-referencing capabilities.

11. Advanced Cost Planning Tools: The system must leverage historical procurement data and AI-driven tools to support IGCE cost planning and decision-making, including cost estimation and comparative analysis features.

12. Enhanced Contract Visibility: The system must improve contract and procurement visibility for all stakeholders through tools such as real-time dashboards or visualization capabilities.

13. Process Visibility and Data Sharing: The system must enhance process visibility and enable seamless data sharing among acquisition team members, incorporating AI-powered workflow tools and collaborative features.

14. Market Research and Vendor Analysis: The system must automate the collection, analysis, and documentation of vendor capabilities using AI tools such as NLP and machine learning.

15. Market Research Trend Documentation: The system must document and track market research trends using tools for automated trend analysis and visualization.

16. Integration of SOW, SOO, and PWS: The system must integrate SOW, SOO, or PWS documents seamlessly within the broader contract file, including metadata tagging and document indexing capabilities.

17. Collaborative Support for Procurement Requests: The system must provide collaborative tools to validate and submit procurement requests effectively, including peer review support, structured validation, and submission tracking.

18. Version Control and Configurable Workflows: The system must implement version control and configurable workflows to support acquisition package development, including tracking changes, managing approvals, and enabling seamless document updates.

19. Iterative and Collaborative Document Reviews: The system must support iterative and collaborative reviews of solicitation documents with features such as real-time editing, comment tracking, and structured feedback loops.

20. Digital Receipt and Tracking of Solicitation Responses: The system must enable the digital receipt, tracking, and management of solicitation responses, including automated tracking, response indexing, and notification features to ensure timely and accurate handling of vendor submissions.

21. Automated Contract Administration: The system must automate contract administration processes, including activity planning, compliance checks, and routine updates, using AI tools to streamline tasks and ensure accuracy.

Data Privacy, Security, and Compliance This category focuses on the vendor's ability to safeguard sensitive procurement and vendor data through robust security measures like encryption and access controls. It also assesses the system's compliance with federal and VA-specific regulations on data protection, cybersecurity, and procurement ethics.

1. Data Security and Privacy: The system must ensure the security and privacy of sensitive procurement and vendor data.

2. Encryption and Access Controls: The system must implement robust data encryption methods and access controls to protect against breaches.

3. Compliance with Regulations: The system must comply with federal regulations such as FAR, VAAR, GDPR, and HIPAA.

4. Role-Based Access: The system must enforce role-based access controls for data and processes to ensure secure access management.

5. Data Integration Risk Mitigation: The system must address data silos and incompatible systems to ensure smooth data integration.

Bias, Fairness, and Transparency This category examines the vendor's strategies for preventing bias in AI models, ensuring that the system promotes fairness in vendor selection and decision-making. It also evaluates the transparency of AI-driven decisions, allowing VA staff to understand and oversee these processes while adhering to ethical standards.

1. Bias Prevention: The system must implement measures to prevent AI bias in vendor selection and decision-making.

2. Transparency: The system must provide transparency in decision-making and enable SAC staff to interpret AI-driven decisions effectively.

3. Ethical Standards: The system must address fairness and ethical concerns in acquisitions to ensure compliance with ethical standards.

4. Collaborative Proposal Compilation: The system must include tools to facilitate the collaborative development and compilation of proposal evaluation reports, including version control and structured validation.

5. Automated Proposal Evaluation Tools: The system must use AI-driven tools to automate the evaluation of proposals, including analyzing technical capabilities, checking compliance with requirements, and ensuring fairness in the evaluation process.

Accuracy, Reliability, and Human Oversight This category evaluates the accuracy and dependability of the AI models, particularly for critical tasks like vendor selection and contract management. It also examines the mechanisms in place for human oversight to intervene in AI-driven processes, especially in high-value or complex contracts, as well as the frequency of model updates and validation.

1. Accuracy and Reliability: The system must ensure that AI models are accurate and reliable for vendor selection and contract management.

2. Human Oversight: The system must provide human oversight mechanisms for AI-driven processes, particularly for high-value or complex contracts.

3. Model Updates and Validation: The system must include frequent updates and robust validation processes to ensure AI model reliability.

4. Document Validation: The system must validate procurement documents to ensure accuracy and completeness.

5. Performance Monitoring: The system must use AI dashboards and predictive analytics for automated performance monitoring.

6. Automated Decision-Support for FAR/VAAR: The system must automate the application of FAR/VAAR clauses during the proposal evaluation phase and provide decision-support tools for ensuring compliance.

7. Integrated Contract Structure Management: The system must manage and integrate contract structures, templates, and execution timelines to ensure compliance and visibility throughout the contract lifecycle.

8. Document Revision Control: The system must ensure proper document version control, including tracking changes, managing revisions, and maintaining a detailed audit trail of edits.

9. Automated Clause Generation: The system must automate the generation of procurement document clauses, ensuring alignment with FAR/VAAR and other legal standards. This must include NLP-based or rule-based tools for this purpose.

10. Contract Modification Support: The system must provide decision-support tools to manage and automate contract modifications, including identifying required changes and ensuring compliance with contract terms.

11. Corrective Action and Non-Conformance Management: The system must include tools to identify, monitor, and recommend corrective actions for addressing contract non-conformance, with AI insights to streamline resolution.

Cost, Implementation, and Vendor Support This category considers the short-term and long-term costs associated with implementing the AI system, including training, maintenance, and updates. It also assesses the level of support the vendor provides for system integration and ongoing maintenance, as well as the measures taken to prevent vendor lock-in to ensure flexibility for future integrations.

1. Cost Breakdown: The system must provide a detailed breakdown of short-term and long-term costs, including training, maintenance, and updates.

2. Vendor Support: The system must include comprehensive support for system integration, training, and ongoing maintenance.

3. Vendor Lock-In Prevention: The system must include measures to prevent vendor lock-in and ensure flexibility for integration with other solutions.

4. Workload Management: The system must provide tools for workload visibility, task tracking, and workforce optimization.

5. AI Incentive Tracking Systems: The system must evaluate and monitor the financial impact of incentives using machine learning or AI models, with features for tracking and reporting outputs.

6. Collaborative IGCE Development: The system must facilitate the collaborative development of IGCEs, including AI-driven decision-support tools for predictive analysis and automated insights.

7. Preliminary Business Case Management: The system must support the creation and management of preliminary business cases, including AI-powered scenario modeling and recommendation tools.

8. Acquisition Strategy Development: The system must provide tools for developing acquisition strategies, including interactive modeling, simulations, and automated decision-support capabilities.

9. Workforce Activity and Resource Visibility: The system must provide tools for real-time visibility into workforce activities, resource allocation, and progress tracking to optimize team performance.

10. Dispute and Appeals Resolution Tools: The system must streamline and support dispute resolution and appeals processes, including case analysis, decision documentation, and automated recommendations for resolution.

11. Complex Contract Negotiation Support: The system must include decision-support tools for managing complex contract negotiations and resolutions, with features such as scenario simulations and AI-driven recommendations.

12. Payment Processing Integration: The system must integrate payment processing tools with decision-support capabilities to ensure accurate tracking, validation, and compliance with payment terms.

Risk Mitigation and Contingency Planning This category reviews the vendor's strategies for mitigating risks such as data integration challenges and handling potential data silos or system incompatibilities. It also evaluates the vendor’s contingency plans for AI system failures or technical issues that could disrupt procurement operations, as well as the system's adaptability to evolving legal and regulatory standards.

1. Risk Mitigation: The system must predict, evaluate, and mitigate contract risks throughout the lifecycle.

2. Contingency Planning: The system must include contingency plans for AI system failures or disruptions to procurement operations.

3. Compliance with Evolving Standards: The system must adapt to evolving legal and regulatory standards to maintain compliance.

4. Fraud Detection and Prevention: The system must monitor and detect anomalies to prevent fraudulent activities.

5. Award Notification Systems: The system must automate the communication of award decisions to stakeholders, ensuring accurate and timely notifications. This must include integration with email or messaging platforms.

6. Scheduling and Coordination Tools: The system must provide capabilities for meeting scheduling and coordination, including preference analysis and integration with existing tools.

7. Automated Procurement Audit and Closeout: The system must automate procurement audit and contract closeout processes, including reconciliation, compliance validation, and accuracy checks, using tools to streamline completion.

Other Considerations Given the nascent and continuously evolving nature of AI capability, please provide information on any additional unique capabilities or practical experience your solution can bring to support the Federal Government procurement/contracting processes.

Does your solution currently have an active or provisional VA Authority to Operate (ATO)? If your solution has an active or provisional, please provide an overview of the ATO, including the scope and what it covers.

Is your AI tool Federal Risk and Authorization Management Program (FedRAMP) authority ready or authorized, have you received FedRAMP authority to operate?

Section 3 – Past Performance VA is seeking a solution that has a proven track record in the U.S. federal space with demonstrated experience supporting Agencies of similar size, scope, and mission complexity as VA. Such experience is critical to ensuring the solution has been proven to be secure and reliable under the required IT security, data management, privacy and other federal regulations that govern information technology implementations within the federal environment.

· Size: Identifies the total dollar value of the contract and the total number of personnel engaged in the effort. A minimum of two (2) relevant experience efforts must meet the minimum size requirement comparable to a National-level effort. A National-level effort is defined as concurrently supporting at least 10 states or 25+ facilities across the U.S.

· Scope: Identifies the aspects of the scope that are the same or like the work requirements as listed in the PWS (i.e., PWS Section ___).

· Complexity: Describes the specialized nature of the engagement to include: a. 1,000+ users under one (1) license; b. 24/7 help desk support; c. Varied staffing levels; d. Provision of subject matter experts; and e. Two (2) or more concurrent contracts/orders.

Please provide the following information regarding your experience deploying your solution for federal customers.

1. Please list all federal agencies where you have successfully deployed your solution in the last five (5) years. For each experience noted, in no more than one paragraph, please provide:

· Agency and Organization name

· Period of Performance

· Solution provided to the Agency.

2. For at least one, but not more than three, of the federal implementations provided in #1 above, please provide the additional detailed information requested below to further demonstrate your relevant past experience.

· Contract and task order number.

· Contract type.

· Dollar value of contract (across full period of performance)

· Period of performance dates

· State whether your company was a prime or subcontractor.

· Customer point of contact (Name, Title, Email, Phone) that VA may reach out to for more information.

· Description of the scope of the work your company performed and its relevance and applicability to demonstrating your solution’s ability to deliver the full set of capabilities outlined in Section 2.

· Additional information where contingent operations were supported by the solution and what objectives needed to be met for contingency operations.

Disclaimer: This Sources Sought Notice (SSN) is for information gathering purposes only as a means to identify interested and capable sources that can provide a solution that fully meets all requirements outlined in Part 2 of this notice. The information provided in this Notice is subject to change and is not binding on the Government. VA has not made a commitment to procure any of the items/services discussed, and release of this SSN should not be construed as such a commitment or as authorization to incur cost for which reimbursement would be required or sought.

VA will not publicly disclose proprietary information obtained as a result of this SSN. To the full extent that it is protected by law and regulations, information identified by a respondent as Proprietary or Confidential will be kept confidential. All submissions become Government property and will not be returned.

Section 4 – Submission All written responses should be submitted in Microsoft compatible formats (e.g., Word) or PDF.

The use of hyperlinks or embedded attachments in RFI responses is prohibited. Any hyperlinks or embedded attachments included will not be reviewed.

RFI Response Sections shall not be submitted in a .zip file.

RFI Sections (Sections 1 thru 4) shall be submitted in separate emails. The format and sections are specified below. Due to VA email file size restrictions (40MB), respondents are encouraged to logically separate their proposal into separate emails. If this is necessary, respondents should attempt to contain a complete section within single email transmissions (and not split sections); however, if this is necessary, please label emails as 1 of 2, for example.

Responses must use 12-point font and 1-inch margins, including all text, tables, and graphics. Additional attachments, brochures, or other marketing materials are not requested and will not be considered as part of the response. Response page limit(s) by section are noted below.

Section 1 – 1 page Section 2 – 15 pages Section 3 – 5 pages Contractors should respond to this RFI by February 19, 2025, by 3 PM ET via e-mail to sandra.rodriguez7@va.gov UNDER NO CIRCUMSTANCES SHALL ANY RESPONDENT CONTACT ANY VA PERSONNEL REGARDING THIS NOTICE.

image1.png