An AI clause suggestion agent reads a draft agreement and works out which clauses should be present for that type of deal, such as audit rights, data protection or termination for convenience.
The agent reads the draft and notes what is missing.
The draft is compared with what similar deals usually include.
Absent clauses or weak wording are listed.
Approved clauses are proposed with the reason.
Legal accepts, edits or rejects each suggestion.
Accepted clauses are tracked in the draft history.
It lists clauses that are missing or thin, and for each one proposes approved wording from your library.
It explains the reason for each suggestion, so the reviewer can accept it or set it aside.
Missing clauses are harder to spot than bad ones, because nothing on the page draws attention to the gap. Reviewers work from memory and personal habits that vary from person to person.
A reviewer reads what is there and may not notice what is not.
A software licence needs different protections from a consulting agreement.
Standard supplier terms favour the supplier and leave out buyer rights.
Two people may ask for different clauses on the same draft.
Business teams press to sign, and a gap may be accepted rather than raised.
The agent identifies the agreement type, value, data handling and term from the draft and the request.
It forms the list of clauses your policy expects for that type of deal.
Each expected clause is checked for presence and strength, and the result is marked present, thin or missing.
For gaps it proposes approved wording from the library and states the policy reason.
Suggestions are shown in order of importance for the reviewer to accept, edit or reject.
Bring a small sample and watch Flo work through it.
| Aspect | By hand | With the agent |
|---|---|---|
| Spotting omissions | Missing clauses are noticed only on careful read-through. | Agent lists clauses a comparable deal would normally carry. |
| Proposing text | Lawyers write the missing clause from scratch. | Approved wording is offered with the reason. |
| Prioritising | All suggestions look equal. | Suggestions are ordered by the risk of omission. |
| Explaining | Reviewers need the why behind each addition. | Each suggestion states the gap it fills. |
| Consistency | Similar deals end up with different protections. | Similar deals draw on the same set of suggestions. |
| If | Then | Who decides |
|---|---|---|
| Draft has no data protection clause but personal data is involved | Suggest a data protection clause | Legal counsel |
| Draft has no termination right | Suggest a termination clause | Legal counsel |
| Clause exists but is one-sided | Suggest balancing wording | Legal counsel |
| Deal type usually carries a service level | Suggest a service level clause | Contract owner |
| Clause already exists | Do not repeat it | None |
| Suggestion conflicts with an existing clause | Show the conflict | Legal counsel |
The agent suggests only the clauses that matter at that size and notes the rest as optional. Legal decides what is proportionate.
It checks referenced documents and does not suggest duplicates. The lawyer confirms the reference is current.
The refusal is shown next to the suggestion with its source. The negotiator decides whether to try again.
The agent states that no comparable norm exists. Legal decides what protections to build.
Talk it through with the team. We will map it to your steps and show where a person stays in control.
Agree which clauses each deal type should usually carry.
Show gaps to lawyers during normal review.
Watch which suggestions lawyers accept and drop the weak ones.
Add further deal types and let buyers see suggestions before legal.
A very low rate means suggestions are noisy. A very high rate may mean reviewers are not applying their own judgement.
Quick answers to what people ask most about clause suggestion.
Book a demoIt is proposing contract clauses that a draft is missing or weak on, using the deal type and company policy. A reviewer decides which ones to request.
It follows the expectations your legal team defines for each deal type and compares the draft against them.
It proposes wording for the reviewer. People make the edits and agree the final text.
No. It narrows the gaps for reviewers to consider, and legal decides what is needed.
Clause library lookup is the task of finding the approved wording for a contract topic, such as limitation of liability or data processing, from a company's maintained clause bank.
→Redline review is the reading of tracked changes a counterparty makes to a draft contract, to understand what has changed and whether each change is acceptable.
→Contract template selection is the step of choosing the correct starting document for a new agreement, such as an NDA, master services agreement or order form, based on what is being bought.
→A fallback clause is pre-approved alternative wording a team may offer when a counterparty will not accept its preferred clause, usually ranked from least to most concessive.
→Bring a small sample of real items. We will show what Flo handles on its own and where it hands over to your team.
