AI grant critique loop

Your grant proposal
has blind spots.
Reviewers will find them.

Simulate a study section before the real one. An adversarial AI attacks your aims, significance, and approach — finding vague claims, coupled aims, and weak preliminary data — then fixes them.

No credit card · Free forever

Three reviewers will read your proposal.
Here's what they'll write.

You submitted the R01

Reviewer 1 wrote: "Significance section is generic — could describe any cardiovascular project. No explanation of why this approach, now." Score: 3.

You submitted the Specific Aims

Reviewer 2 noted: "Aims 2 and 3 depend entirely on Aim 1 succeeding. If Protein X isn't the target, the entire project collapses." Weakness.

You submitted the K award

Reviewer 3 flagged: "Preliminary data shows a trend but n=4. Feasibility is aspirational, not demonstrated." Not discussed.

You knew these were weak when you wrote them. But after months of staring at your proposal, you can't see it the way a fresh reviewer will.

One aims page, four steps

Watch a Specific Aims page go through the loop.

Your text

“Cardiovascular disease remains a leading cause of death worldwide, and new therapeutic approaches are urgently needed. Our lab has extensive experiencein this area and is uniquely positioned to make significant contributions.”

Step 1 · Critic
HIGH

"Leading cause of death" — every cardiovascular grant opens this way. What specific gap in understanding makes YOUR project necessary right now?

HIGH

"Urgently needed" — why? What recent finding or clinical failure creates the urgency? This reads as boilerplate.

MED

"Extensive experience" and "uniquely positioned" — assertion without evidence. What specific preliminary data or expertise makes your team the right one?

Step 2 · Scorer

3 substantive issues found. Score: 1 — proceed to fix.

Step 3 · Fix

“Despite advances in reperfusion therapy, post-MI cardiac remodeling still affects 30% of survivors, with no approved therapies targeting the fibrotic cascade directly. Our 2024 Nature Medicine paper identified Protein X as the switch between adaptive and maladaptive remodeling — the first demonstration of this mechanism in vivo (n=24, p<0.001).”

Same voice · every claim now backed by specific data

Step 4 · Regression check

All original assertions preserved. Cardiovascular disease framing retained, lab positioning maintained. Done.

The loop doesn't invent data. It forces you to replace boilerplate with the specific evidence and rationale that makes YOUR proposal fundable.

See what your aims page looks like after the loop.

No credit card · Free forever

Built for researchers who can't afford a triaged proposal.

PIs writing R01 renewals
Postdocs writing K99/K awards
Faculty writing NSF CAREER grants
Teams writing multi-PI U awards
Researchers writing foundation LOIs
Anyone resubmitting an unfunded proposal
Graduate students writing F31 fellowships

What happened after they ran the loop.

NIH R01

My R01 was triaged twice. Ran the aims page through the loop and it found that my significance section was indistinguishable from any other cardiovascular grant. Made me add the specific clinical gap and our preliminary data. Scored in the 15th percentile on resubmission.

Dr. Michael R.

Associate Professor, Cardiology

NSF grant

The loop caught that my Aim 2 completely depended on Aim 1 working. I'd been staring at it for months and didn't see it. Restructured the aims to be independent with shared methodology. Reviewer comments went from 'fatal flaw' to 'well-designed.'

Dr. Sarah K.

Assistant Professor, Neuroscience

K99 award

I was writing my K99 and my mentor kept saying the approach was 'hand-wavy.' The critic found four places where I said 'we will use state-of-the-art techniques' without naming the technique. Four minutes to get specific, actionable fixes.

Dr. Ananya P.

Postdoctoral fellow, Immunology

Foundation LOI

Submitted a foundation LOI that said our program 'serves hundreds of underserved youth.' The loop pushed me to write '340 students across 12 Title I schools in the 2023-24 year.' Same claim, now fundable. Got the meeting.

Dr. James T.

Director, Community Health Research Center

This is not another AI writing tool.

ChatGPT rewrites your aims

We attack them like a study section reviewer

Grammarly fixes grammar

We fix significance, innovation, and feasibility

Grant consultants take weeks and $5K+

The loop converges in ~3 minutes

Colleagues say 'looks good'

We find what they're too polite to say

How it works

What happens in 3 minutes

0:00

You paste your proposal

Drop in your Specific Aims, Significance section, Approach, or full proposal. Pick your model and hit start.

0:05

Round 1: The first attack

The AI reads your text like a skeptical reviewer — looking for boilerplate significance, coupled aims, vague methodology, and unsupported feasibility claims.

0:20

Real problems only

A separate check filters out nitpicks. Only substantive issues — the kind that get written up in a critique — survive.

0:30

Surgical fixes

Each real problem gets fixed with the smallest possible edit. Your scientific voice, structure, and argument stay intact.

0:45

Nothing was lost

Every data point, citation reference, and scientific claim from your original is re-checked. If a fix accidentally dropped something, it gets flagged.

1:00

Round 2 starts automatically

The improved text gets attacked again. New round, fresh eyes. This catches problems the first round's fixes may have introduced.

2:30

The loop stops itself

When consecutive rounds find only minor issues, or the text barely changes between rounds, the system stops. No wasted rounds.

3:00

Your proposal is reviewer-proof

Every significance claim is specific. Every aim is independent. Every method is named. Same voice — zero boilerplate.

FAQ

The questions you're already thinking.

Why this isn't optional.

You can’t review your own proposal objectively.

After months of writing, every claim feels obvious and every aim feels independent. A fresh adversarial critic evaluates what’s actually on the page — not what’s in your head.

Your colleagues are too polite.

Internal reviewers say “looks strong.” Study sections don’t. The loop gives you the brutal feedback you need before the real review, not after.

One weakness can sink a proposal.

A single coupled aim, a vague significance claim, or insufficient preliminary data can move you from fundable to triaged. The loop systematically finds every weakness, not just the obvious ones.

Resubmissions introduce new problems.

Fixing one reviewer’s concern often creates a new weakness elsewhere. The regression check catches what you miss — every assertion from your original is re-verified after each fix.

Your next submission
has blind spots the study section will find.

Find them first.

No credit card · Free forever · Takes 30 seconds to start