Viktor.comAn AI coworker that actually does the work
ClauseShift: AI Contract Review
Contract review that cites every clause it flags
6 followers
Contract review that cites every clause it flags
6 followers
ClauseShift reviews contracts and shows its receipts. Every risk it flags quotes the exact clause it came from, so you verify the wording yourself instead of trusting a black box. Paste or upload a contract for a plain-English breakdown: risky clauses, what they mean, and a risk score, each linked to the source text. Track deadlines, re-check revised drafts, keep it all private. Free to start, no card. On web, Android, and the Amazon Appstore, with more platforms coming.
How well does it handle heavily redlined drafts where clauses have been edited dozens of times — does the risk score update clause by clause or only at the document level?
@melahat68383 Hey Melahat! Great Question.
Both. Every finding is rated per clause and quotes the exact wording; the 0-10 score is just the rollup. So you can see which clause is driving the risk.
Redlined DOCX? We read it as if all changes were accepted, so you review the current draft, not the markup.
No automatic version diff yet. That's next on the roadmap.
@melahat68383 Quick update: we shipped this today. Run your revised draft as a re-review and the report now shows exactly which findings were resolved, which survived the rewording (with risk changes flagged), and what new risk crept in, plus the score delta. Every finding is still clause-level with the exact wording quoted. You can also now ask the report questions ("can I terminate early?") and it answers only from the contract, citing the clause. Would love to hear how it handles your heavily redlined rounds.
The fact that it links each flagged risk back to the exact clause is genuinely useful, no more hunting through pages trying to figure out why something scored red. Uploading a standard NDA took under a minute and the plain English breakdown actually made sense.
@mer1060139 Thank you, this is exactly the experience we were going for. The whole product bet is that you shouldn't have to take an AI's word for it, if we flag a risk, we owe you the clause it came from so you can check the wording yourself. Glad the NDA breakdown read like plain English too. Since this morning you can also ask the report follow-up questions and re-run a revised draft to see what changed between rounds. If anything scores in a way that doesn't make sense, tell me, that feedback is gold on day one :)