Write better grants

Introduction

A major hurdle for many researchers is translating specialized science into the highly structured language grant funders expect.

We’re going to look at how to use AI to handle that heavy lifting—rubric alignment and administrative cross-referencing. This isn't about the tool doing the thinking; it's about freeing you up to focus on the nuanced components of the grant highlighting your specific areas of expertise.

Today's shift is simple but powerful: we’re moving from "writing into the void" to "writing for the rubric."

We’ll demonstrate using NotebookLM as a "Grant-Alignment Engine." This involves securely uploading dense Requests for Proposals (RFPs) alongside draft abstracts to deliberately identify structural gaps and generate targeted adjustments before submission.

We’re focusing on NotebookLM’s "source-grounding" capability. This means the AI only responds based on the content you’ve uploaded, not information from the broader web. In grant writing, rigorous alignment to funder constraints isn't just a hurdle—it’s the fuel for success.

Instead of exhausting yourself manually cross-referencing a 50-page RFP against your abstract, you can instruct NotebookLM to act as an expert reviewer. By creating this "closed-loop" environment with your specific documents, you get an objective, high-level critique.

Crucially, because NotebookLM grounds its responses only in the uploaded documents, it acts as an administrative partner that evaluates the abstract strictly against the reviewer rubric without bringing in any additional outside information.

Finally, we will demonstrate how this accelerates the grant writing process, ultimately driving institutional impact and student research opportunities.

It is important to note that you should adhere to any requirements or recommendations regarding the use and disclosure of AI in the submission.

Demo

Step 1: Open notebooklm.google.com and create a new notebook.

Step 2: Upload the sources form the linked folder.

Step 3: In the prompt box, enter the following prompt

Act as a meticulous NSF Grant Reviewer, identify exactly where this Draft Abstract fails to address the 'Socio-technical Integration' requirement mentioned on page 8. Provide a bulleted gap analysis and suggest specific terminology from the RFP to strengthen the abstract.

Step 4: Review the output.

The resulting output provides the faculty member with a precise gap analysis, highlighting which criteria are missing, and offers structured, vocabulary-specific suggestions tailored to the agency's language. This is a critical lesson in teaching researchers how to ethically leverage AI for strategic positioning and administrative efficiency.

How does this grounded approach protect the 'intellectual heart' of your research compared to asking a general AI to just 'write a grant'?

Practice

Now, it’s your turn to architect your personal 'Grant-Alignment Engine.' Reflect on a past (unfunded) abstract, a current project draft, or a broader scholarly initiative you wish to fund.

Your task:

1. Open notebooklm.google.com

2. Click the New Notebook button (represented by a plus icon) at the top of the screen to initiate a blank project.

3. Click Add Sources and upload a PDF of a target funding agency’s mission statement or a specific RFP, along with your own draft abstract.

If you don’t have any source documents to hand, you’re welcome to use this Notebook and enter this prompt:

Act as an NSF reviewer. Compare my abstract against the 'Review Criteria' section of the uploaded RFP. Identify where my narrative fails to address 'Broader Impacts.' Give me a 3-point gap analysis and suggest specific vocabulary from the RFP I should integrate to strengthen my proposal."

1. Run your prompt. Review the gap analysis and the suggested vocabulary.

Optional in-person facilitator context for prompting:
Use the P.A.R.T.S. framework to build your prompt:

• P - Persona: 'Act as a meticulous NSF reviewer and strategic communications expert.'

• A - Aim: 'Identify gaps between my abstract and the RFP’s "Broader Impacts" criteria.'

• R - Recipients: 'An interdisciplinary panel of federal reviewers.'

• T - Theme: 'Focus on framing and alignment; do not alter the underlying science.'

• S - Structure: 'A bulleted Gap Analysis followed by three actionable recommendations.'

In Person Peer Review (2 mins):

• Does the recommendation help you better articulate your work without changing the underlying science?

Virtual Reflection (2 mins):

Reflect on your results.

• Does the recommendation help you better articulate your work without changing the underlying science

Reflection

Let’s pause and reflect.

• How might you use this workflow to assist you or other faculty or in your department with their research proposals?

  • How could this workflow be scaled at a departmental level?

Think of AI as a strategic amplifier—like having a consultant on call. It handles the alignment and the framing, but the intellectual heart of the proposal remains entirely yours.

An action item for you is to deploy this workflow for your next grant cycle or departmental research initiative. Demonstrate for your peers how AI can navigate complex administrative requirements and provide formative structural feedback without usurping the role of the principal investigator.