NOAA NESDIS KM_NLP BAA Community Day Presentation.pptx

Broad Agency Announcement (BAA) A Study to Determine Natural Language Processing Capabilities with the NCCF Open Knowledge Mesh (KM/NLP) May 22, 2024 Community Day Briefing

National Environmental Satellite, Data, and Information Service Beau will present

NOAA NESDIS KM/NLP BAA Community Day Wednesday, May 22st (times in EDT) 10:00-11:00 AM: Plenary Session Presentations by NOAA:

Introduction/Administrative and Opening Acquisition Remarks from NOAA Leadership NESDIS Joint Venture Partnership Program overview Overview of Community Day and Ground Rules by Contracting Officer KM/NLP BAA details

11:00 AM: General Questions and Answers Session Please drop questions in the chat box and include organization name

11:55 AM: Closing Remarks

1:00-5:00 PM: 1-on-1 Sessions (20 minutes each)

Thursday, May 23st (times in EDT) 10:00 AM - 12:00 PM and 2:00 - 4:00 PM: 1-on-1 Sessions (20 minutes each)

Community Day Coordination Plenary Session Please sign in using link in chat Please insert your questions in the chat box NOAA team will review questions and provide verbal responses Questions that cannot be answered in Plenary will be taken as actions to respond

One-on-One meetings (Previously scheduled) 20 minute private discussions, every half hour Core Government Evaluation team and AGO will represent NOAA Questions will be recorded Non-proprietary questions that pertain to all potential offerors will be posted publicly Questions that cannot be answered in one-on-ones will be taken as actions to respond

Agenda Welcome & Introduction NOAA/NESDIS/SAE Joint Venture Partnership Program Goals of KM/NLP BAA BAA Process Contacts for Questions Q&A Session General Questions and Answers Closing Remarks

Welcome & Introduction The Joint Venture Partnership was established to leverage capabilities being developed by other federal partners, academia, & industry to provide high return on funds

This program will demonstrate that new technologies will reliably meet NOAA’s operational needs Provide confidence that an emerging technology can be used in an operational environment Reduce risk for new technologies Demonstrate the appropriate reliability

It will support our future architecture and mission performance and allow us to define requirements for future operational procurement

Harshesh will present

Why we do a BAA?

Since NOAA is looking for the best means to meet it mission needs, the BAA allows for all types of implementations and technologies to be submitted for potential study

A BAA allows flexibility when the end result is not yet fully developed and needs agility in the implementation

Harshesh will present

Why is KM/NLP important to NOAA?

NOAA:

has identified a need to improve and expand the open discovery, access, use, reproducibility, and trust of its environmental intelligence.

seeks to facilitate key strategic initiatives, including the advancement of social environmental equity and the enhancement of weather and climate resilience across communities and natural ecosystems.

desires to support a wider diversity of users in the ways they ask to be supported through a single interface.

The envisioned interface will support contextual multi-modal user interaction by integrating services, data, and information from NOAA and non-NOAA sources.

Why is KM/NLP important to NOAA? (2) NOAA is interested in studies that propose solutions integrating proven or nascent advanced artificial intelligence (AI) and machine learning (ML) technologies in the areas of:

Natural language processing (NLP), Networked semantic knowledge graphs (KG), Graph neural networks (GNN), Transformer-based foundation models (FMs) encompassing large language models (LLMs), Knowledge fusion techniques May include prompt engineering approaches such as Retrieval Augmented Generation (RAG), or federated KG query approaches such as virtualization/materialization.

BAA Technical Content This BAA* (located on SAM.gov) is seeking to fund joint partnerships in applying semantic knowledge graph, natural language processing, graph neural network, and foundation model techniques.

The objective of the study will be the generation, assessment, and delivery of processes, software, and information models that demonstrate solutions around one or more areas of knowledge mesh construction, grooming, and exploitation.

Processes described as part of study solutions should be designed in a way that aligns them with the ideals and overview of the BAA, which in part, means:

Broadly leveraging community standards, such as OGC data transfer and format standards, W3C semantic web standards, and targeted semantic metadata vocabulary standards like GCMD, SWEET, Schema.org, Science-on-Schema.org, WORMS, and others, depending on the scientific domains the proposal addresses.

Conforming use of solution within a process/event oriented architecture that decouples microservice computation from information model storage.

Meeting legal licensing requirements that would allow solutions to be leveraged in an open knowledge mesh context. Includes marketplace and role-based access. … *The BAA will be exploratory only and currently there is no planned mission or operations for KM/NLP at this time. The Government does not expect to provide GFE or GFS for this work, except for the items listed in the BAA (i.e. training datasets, existing algorithms, technical and scientific guidance, etc.) which is provided via DEEVA.

Ryan will present

Successful Completion of BAA Goals includes:

Examination and delivery of processes that incorporate NLP, KG, and FM/LLM technologies for the purpose of satisfying stated NOAA strategic objectives, by demonstration and validation of solution through practical scenarios and use cases Examination and delivery of processes and models that provide interfaces conforming to the rules listed above around structure and content.

Examination and delivery of processes and models that demonstrate practical value to knowledge mesh goals so that NOAA may understand how to specifically leverage the outlined technologies to meet stakeholder outcomes.

Delivered solutions for addressing uncertainty in a contextual way and demonstrates ability to convey uncertainty to users on-demand or alongside responses.

Delivered solutions that demonstrate their performance over time and resistance to both data and concept drift.

Trained models to be described and captured by the knowledge mesh as contextual metadata.

Cost models covering cost-per-query, ideally broken down into cost per Task execution and rolled up into cost per Process execution.

Processes and software developed should meet the criteria for portability and transferability.

Documenting the lessons learned from this project and recommendations on how to proceed with further development;

A summary of outcomes at an open public workshop, and the final report includes a version suitable for public dissemination.

Ryan will present

Process and Pricing for the BAA Process:

Solicit Proposals from industry on concepts responsive to technical content of the BAA Proposals (not to exceed 10 pages) reviewed by a technical evaluation board* Selection based on strength of concept and available funding

Pricing:

Offerors will be required to provide detailed pricing for the study per the guidance in Section 3 of the BAA.

*Technical Merit factor is more important than the Past Performance factor. Price is the least important evaluation factor

Mark will present

Partnering & Collaboration

Forming a team with other interested entities to submit a response to this BAA is certainly an option In the case of a team submission, the Government will only interact with the entity identified as the Prime organization. The Prime organization will submit the proposal, and will be responsible for putting in place any sub-contracts and other teaming arrangements.

KM/NLP BAA Timeline Target Schedule Proposal due date – June 12, 2024 Selection of BAA awardees - by end of Aug 2024 Contract Award(s) - by Sept 20, 2024 Contract start date - Oct 1, 2024 Contract completion - NLT Sep 30, 2026

Contracting Officer Introduction Contracting Officers Gabriela Bravo, NOAA Acquisition and Grants Office (AGO) Mark Sullivan, NOAA AGO

Contract Specialists Bryton Curtis, NOAA AGO Lillie Joseph, NOAA AGO

Engagement for these BAAs:

Gaby and Mark, with a cc: to Bryton and Lillie, are the sole points of contact for questions and exchanges of information

QUESTIONS

Contacts for follow up questions and instructions:

Gabriela Bravo gabriela.bravo@noaa.gov

Mark Sullivan mark.a.sullivan@noaa.gov

Bryton Curtis bryton.curtis@noaa.gov

Lillie Joseph lillie.joseph@noaa.gov image6.png image2.png image4.png image1.png image5.png