FDA-SSN-75F40125Q00192_RAPID VI_Instructions.pdf

SOURCES SOUGHT NOTICE (SSN)

FDA-SSN-75F40125Q00192

FOR: Real-World Application for Innovation and Development (RAPID) Sources Sought

DEPT/AGENCY INFORMATION: Department of Health and Human Services U.S. Food and Drug Administration (FDA)

Requiring Offices: Center for Drugs Evaluation and Research (CDER)/

Office of Translational Sciences /Office of Computational Science

(OCS)

TYPE OF NOTICE

The U.S. Food and Drug Administration (FDA) is issuing this sources sought notice as a means of conducting market research, pursuant to FAR Part 10 Market Research, to identify potential parties from the marketplace capable of providing services for CDER/OCS’ Real-World Application for Innovation and Development platform.

This is a Sources Sought Notice ONLY, not a Request for Quote (RFQ) and it does not commit the Government to award a contract now or in the future. The purpose of this sources sought is to identify organizations, in particular SMALL BUSINESSES on GSA MAS Schedules 54151S Information Technology Professional Services and 54151HEAL Health Information Technology Services, with the capabilities to provide the services stated herein.

Submission of any information in response to this market survey is purely voluntary and the FDA welcomes responses from all interested parties; the Government assumes no financial responsibility for any costs incurred.

The FDA does not intend to make a vendor selection decision or award a contract on the basis of responses received. Acknowledgement of receipt of responses will not be made, nor will respondents be notified of the outcome of the FDA's review of the information received.

Additionally, the FDA does not intend to hold discussions concerning this notice with any interested parties. However, FDA reserves the right to contact vendors if additional information is required.

BACKGROUND

The U.S. Food and Drug Administration (FDA) is issuing this source sought notice as a means of conducting market research, pursuant to FAR Part 10 Market Research, to identify viable industry parties capable of supporting FDA’s Real-World Application for Innovation and Development (RAPID), enhancement of the current AWS platform, development of new applications prototypes and Proof of Concepts.

The mission of CDER’s Office of Computational Science (OCS) is to modernize the Center’s regulatory review process and enable reviewers to spend less time dealing with complex data and more time applying their expertise to information through the use of integrated tools, services, and training. OCS’ efforts to accomplish this mission involve developing technology and analytics to support regulatory review work. RAPID was developed as a means to implement a cloud-based and on-premises environment to develop and continuously improve the applications and databases needed to meet the innovation and analytical needs of FDA users.

The RAPID platform is capable of supporting a variety of cloud and other operations including, supporting individual regulatory research, ingest, evaluate, and validate new analytic tools and methodologies, support new methods of multi data base curation, characterization and integration for transdisciplinary analysis of complex and big data, and a range of artificial intelligence tools and use cases for FDA needs.

RAPID is a sandbox that promotes the development of innovative analytic tools, supports data science research, and the evaluation of cutting-edge technologies. The RAPID System in its current form uses Artificial Intelligence (AI) and cloud-based compute facilities to develop innovative analytical solutions that support regulatory decision making and scientific analysis.

Refer to the National Academies of Sciences “Convergence”, https://www.nae.edu/113283.aspx, report to get an understanding of what ”transdisciplinary FDA scientific need” must include.

OBJECTIVE

Enable the RAPID) Digital AI Platform program to design, develop, deploy prototypes and proof-of-concepts to meet FDA’s mission critical needs. Expand the RAPID platform’s capability building upon the lessons learned and functionality provided by RAPID using innovative architecture to support:

• Continue supporting and maintaining the development of existing databases, technologies, and applications in the RAPID Environment. Provide maintenance support for the FDA RAPID environment and deployed applications in the RAPID environment and support the deployment and enhancement of tools and applications into FDA’s production environments. (i.e., On premise, Cloud Platforms).

• Conduct system upgrades for deployed applications to continually improve the ability for the FDA to meet their mission critical real-time data management and analytical capabilities.

https://www.nae.edu/113283.aspx

• Continue to establish and maintain the exploration/experimentation sandbox that supports innovation and analytics development in support of regulatory review the development of existing databases, technologies, and applications in the RAPID Environment.

• The FDA expects accelerated advancements in various AI and Cloud technologies, both software and hardware, over the next five (5) years and are seeking support in enabling the RAPID program to be at the forefront of these advancements.

The support includes, but is not limited to:

• Human Centered Design (HCD) which accelerate design into deployment

• Generative AI and AGI

• Data Cataloging and Data Integration solutions

• Rapid design and development of new use cases

• Seamless Multi-Cloud Solutions

• Data Mesh Solutions

• Analytical Tool Development

• Serverless Analytics

• Mission AI

• Real-Time Clinical Trials Frameworks

• Explore data integration solutions

• Software development

• Exploration of AI/ML

• Natural Language Processing (NLP)

• Large Language Models (LLM)

• Analytic tool development.

CURRENT RAPID STATE

The RAPID system was developed originally to provide a bi-directional information transmission pathway for drug safety information and provide to FDA scientists real time data analysis system. The design requirements for RAPID were identified by referencing data and analytic problems which occurred during the drug safety evaluation of Peramivir, an unapproved drug publicly released under the Emergency Use Authorization, (section 564 of the Federal Food, Drug, and Cosmetic Act (FD&C Act)), and used as salvage therapy for severely ill patients in during the H1N1 epidemic of 2009-2010. Since 2013, RAPID was developed to meet CDER requirements of performing drug safety surveillance and analysis during Medical Counter- Measures (MCM) events like the H1N1 epidemic.

The RAPID system has evolved into an AI and Analytics platform for tools supporting innovation and analytics for regulatory review processes. RAPID is FDA CDER’s Innovation As-A-Service (IAAS) Platform, which provides experimentation sandboxes and a production environment in a multi-cloud environment. RAPID is designed to explore data sets, develop and test new analytics, models, and technologies. RAPID minimizes the barriers to designing analytical solutions by merging data, infrastructure, development support, artificial intelligence, and knowledge management.

RAPID cloud allows safety evaluators and medical officers access to modern technologies to make critical review decisions and enables FDA reviewers and collaborators to address product safety issues and improve patient care. The RAPID Data and Analytics Hub supports innovation efforts in CDER and provides a space or technologists for innovators to prototype and advance new approaches, established RAPID Data Hub with Data connections to data sources and managed Platform Service. The space is highly used as a prototyping environment to explore new technologies and generate operational data to help minimize the cost and risk of operationalizing innovative analytics and infrastructure on an enterprise scale for CDER.

FUTURE STATE:

The future of RAPID is in leading the adoption of AI and leading-edge technologies to support FDA use cases across heterogeneous cloud environments.

RAPID needs to continue to adapt to the changing technology landscape and provide the FDA with the next generation of tools, technologies, models, and approaches to ensure the FDA can meet critical mission needs.

While AI technologies are now available to the general public, the development of targeted AI solutions that leverage or use novel techniques or data still require large investments in both compute and data processing. The leading AI companies, such as Google, Microsoft, OpenAI, NVIDIA, etc spend billions of dollars a year in both computing and collecting data for training their models. To build and develop leading targeted AI solutions that meet FDA use cases, the FDA requires contracting services partners that can work effectively with large, medium, and small organizations to design and develop new solutions that meet FDA requirements. In addition, the skills required to develop and scale novel AI solutions or models require a diverse set of skills and resources, such as Software Engineering, AI model development, Cloud Computing Architecture, DevSecOps, and AI Risk Management.

The future of cloud computing at the FDA will most likely be a heterogenous cloud computing environment with multiple Cloud Service Providers (CSPs) such as Amazon AWS, Microsoft Azure, Google GCP, Salesforce, SnowFlake, etc providing different services to the FDA. Using and deploying solutions, with advanced AI approaches, across heterogenous cloud environments will require the FDA to have contracting services support that understand different CSP environments and how to design, build, and deploy solutions that scale to meet FDA use cases.

RAPID Capabilities & Functions

The RAPID platform enables the FDA to meet its mission critical real-time data management requirements during public health events that require the timely access to data and tools, processing, analytics, and remediation that used to take months, weeks, or days must be redefined and reduced to minutes and seconds. Information must be collected, processed, and analyzed when and where needed.

• Support and development of the research exploration/experimentation sandbox that supports innovation

• Enhance and maintain the project intake and project governance structure for scientific projects in the RAPID space.

• Enhance and maintain a searchable system for knowledge management for all projects performed in the RAPID space.

• Enhance, maintain, and manage licenses and applications that supports innovation by FDA scientists and by the Contractor for several analytics packages.

• Develop and enhance FDA’s adoption and use of artificial intelligence (AI) and machine learning (ML) in analyzing regulatory review data.

• Maintain the data hub and integration of data in the RAPID environment.

• Provide extract, transform, and load (ETL) services for projects supported in RAPID.

RAPID CAPABILITIES

The FDA seeks to enhance existing RAPID capabilities by building upon the lessons learned and functionality provided by the RAPID prototype using innovative architecture to support Analytical tool development (including, but not limited to):

Analytic Tool Development.

R-Studio Connect Jupyter Analytics Qlik Sense Tableau Analysis Studio

RAPID USE CASES

Examples RAPID Use Cases include, but are not limited to the following:

Development of an App Support Teams Biopharmaceutics automation tools, which are custom coded tools with the goal of increasing the efficiency and accuracy of dissolution similarity calculations for NDA and ANDA Original Applications, as well as SUPAC submissions.

Designing the operational environment to scale multi cloud/multi-modal AI use cases and create consistent design pattern that can be leverage across the FDA to meet mission needs.

Using Large Language Models (LLM) and other python tools to process Establishment Inspection Reports (EIR) to allow users to upload EIRs and save outputs in various formats.

Analyzing FDA Adverse Events Reporting Site (FAERS) narratives to extract adverse event data from unstructured text content

Supporting risk optimization thresholds analysis to help characterize and prioritize commodity risks.

Supporting R-Shiny web applications that extracts information from documents for business decision making and returns.

Utilize Artificial Intelligence methodologies to assist FDA stakeholders and internal scientific team to find and track regulatory information.

Explore Hugging Face proxy servers and GPU clustering as a scalable solution for generative AI in FDA.

Evaluating natural language processing (NLP) tools that improve the efficiency and consistency of writing drug deficiencies.

The creation of new data sources and data architectures to accelerate FDA ability to solve complex regulatory problems.

Develop, enhance, and provide technical support for novel prototype NLP machine learning and data analytics tools capable of extracting safety and efficacy information from unstructured free text from various technical documents and reports, including but not limited to FDA-approved product labels, published scientific literature reports, clinical narratives from spontaneous adverse event reports and deidentified electronic clinical trials data.

Exploring innovative AWS based analytical approaches to evaluate their ability to address CDER analytical needs in line with the current vision for IT architecture

Maintain a transparent RAPID website to make available technical information to users on available services in the RAPID space with additional information regarding project details and their status.

Provide O&M technical support for each tool and application, including troubleshooting to resolve complex problems, identifying, and resolving system errors, supporting software installation and configuration, incorporating patches and updates, computing platforms and environments.

GENERAL SUBMISSION INFORMATION

A complete response to the questions below should clearly state respondent’s capability to support RAPID capabilities and functions described herein and in the draft statement of work.

Please note that a respondent's failure to provide a complete response to a question may be interpreted to mean that it does not understand the question and does not demonstrate the capability to provide services required. Responses should demonstrate capability not merely affirm the respondent's capability (i.e. the response must go beyond the statement that, "XYZ company can provide technical support").

Respondents will not receive individualized feedback on any suggestions. No basis for claims against the United States government will arise as a result of a response to the questions within this notice or from the United States government's use of such information. Proprietary, classified, confidential, or sensitive information should not be included in your response. The Government reserves the right to use any non-proprietary technical information in any resultant solicitation(s).

Responses should be emailed to Noah Padilla, Contract Specialist, at Noah.Padilla@fda.hhs.gov by 12:00 PM ET, May 27, 2025.

Please include the number “FDA-SSN-75F40125Q00192: Real-World Application for Innovation and Development (RAPID)” in the subject line. No phone calls will be accepted for this notice.

CAPABILITY STATEMENTS SUBMISSION

Interested, capable vendors (parties) will submit contact information and a Capability Statement that shows evidence of their understanding of the overall RAPID need and capabilities in the following areas:

1) Contact information:

Company Name:

Mailing Address:

Name, telephone number, and e-mail address of a point of contact having the authority and knowledge to clarify responses with Government representatives:

UEI:

Socio Economic Status:

Tax Identification Number, and company structure (Corporation, LLC, partnership, joint venture, etc.). Companies also must be registered in the SAM (www.sam.gov) to be considered as potential sources:

Your proposed NAICS for this effort:

Your proposed PSC for this effort:

Identify Governmentwide Acquisition Contract(s) name and number (i.e., GSA, NITAAC) that can be leveraged for these services.

All text in the body shall be black. Interested parties may choose to use color in graphics, charts, or tables for greater clarity. Text in graphics and charts shall be no smaller than 9-point Arial Narrow, and text in tables shall be no smaller than 9-point Arial. No fonts may be condensed.

mailto:Noah.Padilla@fda.hhs.gov

The response is limited to ten (10) pages. Quoters shall use only 8.5 inch by 11-inch pages, single-spaced pages with margins no less than one inch on each border and shall number each page consecutively. FDA prefers that Interested parties use Arial 11 font, but they may use any legible font that is equal in size to or larger than Arial 11.

Note: A page that contains a table, chart, graph, etc., not otherwise excluded below, is subject to the page limitation.

The following items are not included in the ten (10)-page limit:

•Cover Sheet;

•Table of Contents;

•Dividers

All critical information from appendices should be summarized in the response.

2) Please provide a response to the following questions in regard to the capabilities of your company:

Innovation:

o Briefly describe your experience with testing applications and scenarios to learn strategies for understanding analyzing and simplifying complex data to support regulatory decision making.

o Briefly describe your experience developing deep learning/artificial intelligence capabilities to generate predictive models related to (but not limited to) toxicities related to marketed human drugs and therapeutic biologic products as well as drugs and therapeutic biologic products under development?

o Briefly describe your experience using innovative analytical development projects using Agile delivery methodologies.

o Briefly describe your experience with Robotic Process Automation and provide an example.

o Briefly describe your experience with language versions (such as a multilingual interface to accommodate MCM reporters for which English is not their primary language).

o Briefly describe your capabilities in building and deploying customized Large Language Models and leading-edge AI solutions.

o Briefly describe your capabilities in building, deploying, and scaling multi-Modal AI across multi cloud environments.

o Briefly describe your capabilities creating novel data sources to address or solve public health challenges.

o Briefly describe your abilities to partner with providers of AI solutions.

Production Development:

o Do you have experience implementing applications into multi cloud environments?

If so, please provide two (2) to three (3) examples with a brief explanation on the deployment.

o Do you have experience developing applications using Windows, LINUX, and a combination of Windows/Linux environments? Provide brief explanation.

o Are you able to provide functionality to integrate outputs from the prototype NLP-based technology and data analytics tool with outputs from existing data mining and analytical tools in FDA for harnessing scientific literature reports, spontaneous adverse event report data, and clinical trials data? Provide brief explanation.

o Do you have the capability to develop an App Support Team? If so, briefly describe.

Operation and Maintenance:

o Do you have experience managing multiple Applications/systems that is integrated between the cloud and on-prem environments? Provide brief explanation.

o Briefly explain your ability to recruit and retain subject matter experts required to do the described work.

o Do you have experience supporting multiple concurrent cloud-based deployments in the Cloud? If so, please provide an example with a brief explanation?

o Do you have expertise provide customer support services to applications/systems in production? Provide brief explanation.

Please provide a YES/NO response for the following questions:

QUESTION

RESPONSE

(Yes/No)

1) Do you have experience with cloud-based Data Storage platforms?

2) Do you have experience with cloud-based GPU Accelerated computing?

3) Do you have any experience with Machine Learning?

4) Do you have any experience with Text Mining?

5) Do you have any experience with Data Science and Machine Intelligence Methods and Strategy?

6) Do you have experience performing FISMA, ATO and 508 compliance in the Federal Government environment?

7) Do you have any experience with Data Acquisition and Curation?

8) Do you have any experience deploying large and complex as well as small scale systems into multi-Cloud environments?

9) Do you have any experience developing strategies for IT transformation?

10) Do you have experience with developing, deploying, and scaling multi-modal, artificial intelligence, and deep learning technologies and techniques across multiple cloud platforms?

3) Address any questions, comments, assumptions that you have pertaining to this notice.

Questions will not be answered via this notice. They are intended to assist FDA with developing its requirement.