Review the essay you have to defend
Quick answer
Nudge Buddy reviews your ENEM essay with questions on the five competencies and flags of what sounds like AI — it reads and asks, never writes for you; it's practice before the exam, not the official grade, and it promises no score.
Paste your draft. Nudge Buddy asks the questions your reader would and flags the phrases that sound machine-made — and you rewrite in your own words. It never writes, rewrites, or detects. The essay stays provably yours.
No card · 5 reads a day · your text never trains a model
Brazil's ENEM essay is worth up to 1,000 points and, on its own, decides a lot of admissions. The trouble is that almost everyone reviews their own draft in the weakest way possible: rereading it hunting for a stray comma. A marker does not stall on a comma. A marker stalls when the thesis never surfaces, when the argument does not hold, when the intervention proposal forgets the agent or the action. Those are exactly the gaps the writer misses — because the writer already knows what they meant to say.
Nudge Buddy is built for that blind spot. You paste your draft and it hands back the questions a marker would ask of each paragraph, mapped to the five ENEM competencies. It does not “correct” in the sense of rewriting: it asks and points, and you rewrite in your own words. You cannot bring Nudge Buddy into the room, but you can walk in having trained the eye the marker uses. And because it writes none of your words, the essay stays provably yours.
To be clear from the start: Nudge Buddy does not give an official score and does not promise a “1000”. No one honest does. What it does is show you, before the marker, where the marker would stop — so you can fix it while there is still time.
What the read looks at
- A clear, defensible thesis
- Evidence that actually supports the claim
- Logical flow from one paragraph to the next
- Cohesion — transitions that connect, not pad
- A conclusion that lands rather than repeats
The ENEM read — both lenses on the five competencies — is free, 5 times a day. Pro runs the same read on our most advanced model and unlocks more reads a day.
See Pro and plans →One example, competency by competency
A static example to illustrate the read. The live demo remains just below.
Prompt (hypothetical)
The challenges of countering disinformation in Brazilian society.
Draft excerpt
The 1988 Federal Constitution guarantees everyone the right to information, yet Brazilian reality contradicts this principle. In recent years, disinformation has spread across social media and compromised important decisions by the public, especially on public-health topics. This happens because digital platforms prioritize engagement over truth, which lets false news reach millions of people in a short time. Therefore, it is evident that fighting disinformation is one of the country's greatest challenges.
What the read sees
- C1 Command of the standard written norm — Punctuation and agreement are under control here; the risk is the generic wording (“important decisions”, “in a short time”) that weakens the line without breaking the rule.
- C2 Grasp of the prompt and structure — The theme is engaged, but the thesis only appears in the last sentence; it lacks a framing that announces where the argument will go.
- C3 Arguments and repertoire — There is an attempt at repertoire (the 1988 Constitution), but it stays decorative — cited and dropped, without tying the right to information to the mechanism of disinformation.
- C4 Cohesion and linguistic mechanisms — The connectives are there (“yet”, “because”, “therefore”), but “Therefore” closes an idea that has not been demonstrated — the conclusion arrives before the proof.
- C5 Intervention proposal — There is none in this excerpt yet; the flag is to reserve room for agent, action, means, purpose, and detailing, respecting human rights.
The Deep Read questions, tied to lines of the excerpt
- You claim disinformation “compromised important decisions by the public” — which concrete episode or figure supports that? Without one, the line is opinion.
- You say platforms “prioritize engagement over truth” — how do you know? Is there a source, an example, or a mechanism the reader can follow?
- The thesis only shows up at the end of the paragraph. Where is your thesis in the opening, and does it signal the two or three lines of argument you will defend?
- You open with the 1988 Constitution and never return to it. How does that repertoire connect to the argument — or is it there only as filler?
One line from Human Voice
Phrases like “it is evident that” and “in recent years” are crutches that show up a lot in machine-made text; Human Voice flags them so you can swap in something of your own — and the swap is yours, Nudge Buddy does not write the replacement.
How it works, step by step
- 1
Paste the draft
Pasted or typed, no required upload. Pick the ENEM mode.
- 2
Read the Deep Read
You get a 0–100 score (an internal gauge, not the INEP grade) and the questions a marker would ask, organized by the five competencies.
- 3
Run Human Voice
The second lens flags the phrases that sound machine-made — triads, padding, empty transitions. It points; it does not rewrite.
- 4
Rewrite in your own words
You answer the questions inside your own text. Every answer is yours — that is what keeps the essay provably yours and defensible.
- 5
Read it again
Run it once more and see what still stalls. That is 5 reads a day on the free plan — enough to polish a whole essay.
See what it looks like on a real draft:
Every city already owns a scarce, valuable asset it gives away for free: the curb. Parking is priced as if land were infinite, and the result is predictable — drivers circling, buses stuck behind them, deliveries double-parked in the bike lane. In this rapidly evolving urban landscape, the fix is not more asphalt but a price that moves with demand. San Francisco tried it and the circling dropped. Ultimately, pricing the curb is not just about parking — it is about fairness, efficiency, and the future of our cities.
Strong instinct, thin evidence. The argument names a mechanism but never shows it working — and the closing paragraph slips into borrowed cadence.
Common questions
Does this turn my essay into a “perfect score”?
No, and no one honest guarantees that. Nudge Buddy gives no official grade and promises no result. It shows you, before the marker, where the marker would stall — off-topic drift, thin argument, incomplete intervention proposal — so you can fix it in time.
How many lines should an ENEM essay have?
It needs at least 7 lines to be graded; the official sheet allows up to 30. What decides the grade is not filling lines but a clear thesis, a supported argument, and a complete intervention proposal. The read helps with exactly that content, not the line count.
Can I use it on exam day?
No. The ENEM is taken offline, with no tools of any kind. Nudge Buddy is for practice, in the days and weeks before — so you walk in having rehearsed the marker's eye. Bringing any aid into the exam is fraud, and it is not what we propose.
Is it reliable?
The read follows the five public ENEM competencies and asks questions within them. We invent no grade and promise no admission. Honesty is the product: Nudge Buddy names the problem and hands the pen back — you are the one who judges the text in the end.
Do you grade like INEP?
No. Only INEP assigns the official grade, with two markers and its own criteria. Nudge Buddy is not the board and does not replace the official correction. It anticipates the questions a marker would ask, so they do not catch you off guard.
Does this count as cheating?
No. Nudge Buddy writes none of your words. Cheating is letting an AI draft the text and handing it in as your own — the opposite of what it does. Here it only asks and points; the text stays entirely yours.
What about Turnitin? Could my text be flagged as AI?
Nudge Buddy is not an AI detector and controls neither Turnitin nor any other detector — that is on them. What we guarantee is the mechanism: we add no words to your text, so there is nothing of ours to flag. We promise no outcome on any detector.
Do I need an account?
To keep a history and use the 5 reads a day, just create a free account — no card.
Is it really free?
Yes. The ENEM mode, with both lenses (Deep Read and Human Voice), is free: 5 reads a day, no card. Pro runs the same read on our most advanced model and unlocks more reads a day. Pro is currently waitlist-only.
Does my text become AI training data?
No. Your text is never used to train a model. It is yours, start to finish.
Does Nudge Buddy write my essay for me?
No, never. It reads and asks, never writes for you — it asks the questions a marker would and flags what sounds like AI, and you do the rewriting, in your own words. The essay stays provably yours.
Is Nudge Buddy an AI detector?
No. Nudge Buddy is not a detector — it gives no “AI probability” and scans your text against no database. Human Voice flags the phrases that sound machine-made so you can rewrite them in your own voice. That’s coaching, not detection.
Will I be penalized for using AI on my ENEM essay?
What typically gets penalized is handing in text you didn’t write. Nudge Buddy writes none of your words: it reads and asks, never writes for you, so there’s no machine text to penalize. And it’s for practice only — on exam day you use no tools at all. Rules on AI use are set by each school or board; follow theirs.
ENEM writing guides
Paste your essay and see where you'd stall
Get your first readNo card · 5 reads a day · your text never trains a model