2  Apogee

AI/ML-Powered Grant Evaluation, Education, and Efficiency Ecosystem

Apogee aims to develop an AI/ML-powered grant evaluation, education, and efficiency ecosystem. The vision for the Apogee encompasses the entire grant lifecycle, from pre-award ideation to post-award management and reporting. By leveraging cutting-edge AI and machine learning technologies, Apogee seeks to augment the support already afforded to researchers throughout the grant cycle.

2.1 Goals

  1. provide researchers tools for pre-grant ideation;
  2. empower researchers with comprehensive, objective, and timely feedback on draft NIH grant proposals that augments traditional mock study section reviews;
  3. support the “uncreative” aspects of the grant writing and submission processes through judicious use of AI/ML tools;
  4. facilitate post-award management and reporting where possible;
  5. promote the dissemination of research results and the development of new ideas and proposals.

2.2 Development Phases

Phase 1

Phase 1 will serve as the so-called minimum viable product1. This phase will deliver a basic AI/ML-powered grant evaluation system that provides researchers with narrative and quantitative feedback on draft NIH grant proposals, supplementing traditional mock study sections and grant-writing workshops.

1 The version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort (similar to a pilot experiment).

This system will leverage natural language processing (NLP) and machine learning (ML) techniques to evaluate writing style, alignment with the RFA, clarity of potential impact, and other key aspects of the proposal. The system will also provide suggestions for improvement and identify areas of strength and weakness.

The Phase 1 system will be designed to accept specific aims pages and the Significance and Impact sections of the grant as well as the RFA to which the grant is responding. It will generate narrative summaries, readability scores, similarity metrics to funded grants, keyword analyses, pointers to related research. The system will be accessible via a web interface and will be integrated with NIH Reporter and PubMed (which now includes preprints from bioRxiv and medRxiv).

Phases 2 and beyond

Based on feedback from Phase 1, subsequent phases will expand the capabilities of the system to cover the entire grant lifecycle. This will include tools for pre-grant ideation, post-award management, and reporting. The system will be integrated with other research tools and databases to provide a comprehensive ecosystem for researchers, grant teams, and administrators.

2.3 Key Considerations

Work on Apogee will be guided by the following key considerations:

  • Security and Privacy: Ensure that all data is handled securely and that researchers’ intellectual property is protected.
  • Ethical Considerations: Address potential biases in the AI/ML models and ensure that the system is used responsibly.
  • User Experience: Design an intuitive and user-friendly interface that researchers can easily navigate.
  • Stakeholder Engagement: Involve researchers, grant administrators, and other stakeholders in the development process to ensure that the system meets their needs.
  • Sustainability: Develop a plan for maintaining and updating the system over time to keep it relevant and effective.
  • Reporting and Transparency: Develop and maintain metrics on the cost, usage, and effectiveness of the system to ensure accountability and transparency.