GuidesPath

Grant deadline sprint

Move from a blank specific-aims page to a submission-ready one-pager and defensible budget justification without losing your voice.

Intermediate
Learning path

Grant deadline sprint

PhD students and postdocs with a fixed grant deadline and no time to babysit AI output. · Move from a blank specific-aims page to a submission-ready one-pager and defensible budget justification without losing your voice.

Lessons5Time~140 minArtifactA submission-ready Specific Aims one-pager plus a budget justification with line-by-line rationale.Levelintermediate
Before you start
  • A confirmed funder, call, and page limit
  • A working bibliography in Zotero or similar
  • Comfort drafting in your own voice — AI is a co-editor here, not the author
Tools in this guide
0%0/5 steps
Path artifact

A submission-ready Specific Aims one-pager plus a budget justification with line-by-line rationale.

Finish every lesson and you should have: A submission-ready Specific Aims one-pager plus a budget justification with line-by-line rationale..

Lock the call before you draft

20 minClaude

Pull the funder's review criteria, page limit, and formatting rules into a single working doc. Highlight the exact phrases reviewers grade against.

Pitfall
Letting Claude paraphrase the call instead of pasting the verbatim criteria.
Verification
Every aim and budget line can be mapped back to a named review criterion.
→ See playbook step 1 in detail

Draft Specific Aims in your voice

35 minClaude

Write three aims yourself, one sentence each. Use Claude only to compress, not to invent.

Pitfall
Accepting the model's polished prose when it has quietly changed your hypothesis.
Verification
A peer can read the aims aloud and tell which one is the keystone.
→ See playbook step 2 in detail

Pressure-test with grounded literature

25 minNotebookLM

Load your bibliography PDFs into NotebookLM and ask which aim is weakest and which prior work most threatens novelty.

Pitfall
Treating a grounded summary as a substitute for reading the threatening paper.
Verification
Each aim has one named threat and one defensible response in your notes.
→ See playbook step 3 in detail

Build a budget with line-level justification

40 minOverleaf

Draft the budget in Overleaf or a spreadsheet, then ask ChatGPT to surface unexplained line items and missing fringe/indirect math.

Pitfall
Letting the model invent indirect-cost rates instead of pulling them from your office of research.
Verification
Every line has a sentence saying why this amount, this year, this person.
→ See playbook step 4 in detail

Reviewer red-team in a final read-through

20 minClaude

Use Claude to roleplay a hostile reviewer against the one-pager and budget; capture every objection that lands.

Pitfall
Asking the model to be 'helpful' instead of 'mean'. Helpful reviewers don't exist.
Verification
Every objection is either addressed in the doc or logged in a 'will not change' note.
→ See playbook step 5 in detail
Finish line

You should now have: A submission-ready Specific Aims one-pager plus a budget justification with line-by-line rationale..

  • Every aim and budget line can be mapped back to a named review criterion.
  • A peer can read the aims aloud and tell which one is the keystone.
  • Each aim has one named threat and one defensible response in your notes.
  • Every line has a sentence saying why this amount, this year, this person.
  • Every objection is either addressed in the doc or logged in a 'will not change' note.