A matter review system for a disputes team

Consider a 12-person disputes team working across six active commercial matters.

Each matter has a partner, one senior associate, two or three juniors, and shared support from paralegals and operations. The team already has document management, email, calendars, time records, task lists, and matter notes. The problem is not that nobody has information. The problem is that the matter review meeting depends on people rebuilding context from too many places.

Every Friday, the partner wants to know what changed, which deadlines are moving, which evidence is still missing, what client commitments are open, and what needs judgment before the next step. A generic AI assistant can summarise documents, but that is not enough. The team needs infrastructure around the review rhythm.

This is where Proximity would be tailored as a matter review system.

The purpose is not to automate legal judgment. The purpose is to make the review meeting better prepared, more source-grounded, and less dependent on whoever had time to assemble the pack.

The Workflow

The workflow is weekly matter review for active disputes.

The team wants a repeatable view of each matter:

• What changed since the last review.
• Which deadlines, filings, and client commitments are coming up.
• Which documents, correspondence, or notes support the current position.
• Which issues are waiting on evidence, instructions, or internal review.
• Which drafts are ready for human legal review.
• Which follow-ups are late or unclear.

That sounds simple, but disputes work rarely sits in one system. A key point may live in an email. A procedural deadline may be on a calendar. A client instruction may be in a call note. A witness point may be in a draft chronology. A concern may have been raised in chat and never added to the matter note.

Proximity would not replace the matter record. It would connect the operating context around the matter so the team can prepare and review it with less friction.

What Proximity Models

For this deployment, Proximity would model the matter as the core container.

The system would connect approved sources such as:

• Document management records.
• Matter notes and chronologies.
• Relevant email threads.
• Calendar deadlines and review dates.
• Task owners and follow-up commitments.
• Client instructions and meeting notes.
• Drafts awaiting review.

The useful model is not just "all text about the matter." It is structured operating context:

• Matter, client, counterparty, stage, and responsible lawyers.
• Source document, version, date, and confidentiality status.
• Open issue, owner, source evidence, and next review date.
• Deadline, basis, owner, and confidence level.
• Draft, reviewer, review status, and unresolved questions.
• Client commitment, source, owner, and due date.

This matters because legal work is evidence-heavy and responsibility-heavy. The American Bar Association's Model Rule 1.1 describes competent representation as requiring the legal knowledge, skill, thoroughness, and preparation reasonably necessary for the work. ABA Formal Opinion 512 on generative AI also stresses duties around competence, confidentiality, communication, supervision, candor, and fees. Those duties are performed by lawyers, not by software. Software can only help if it makes preparation and supervision easier to perform.

What The System Prepares

For each weekly review, Proximity could prepare a matter review packet.

The packet would include:

• A change summary since the last review, with source links.
• A deadline and commitment list, with owner and evidence.
• A missing-context section for facts, documents, approvals, or instructions that are not yet clear.
• A draft chronology update, explicitly marked as draft.
• A review queue for documents or correspondence that need a lawyer's attention.
• A follow-up list showing what was promised, who owns it, and what evidence supports it.

The system could also prepare briefing notes for the partner before the meeting. Those notes should separate confirmed facts from open questions and assumptions. They should show where each point came from. If a point is not supported by the approved matter record, it should be flagged as missing or uncertain.

This is source grounding applied to matter review. The output is useful because it reduces search time and exposes gaps. It is not useful because the model has its own legal view.

What Remains Human

The legal opinion remains human.

The system must not decide litigation strategy, advise the client, approve correspondence, determine privilege, file documents, make legal arguments, or decide whether evidence is sufficient. It should not send client communication or update the official matter position without human review.

The partner and lawyers remain responsible for:

• Legal analysis.
• Professional judgment.
• Strategy.
• Client advice.
• Final wording.
• Privilege and confidentiality decisions.
• Approvals and filings.
• Any external communication.

Proximity's role is infrastructure. It prepares the room before the professional work happens.

That boundary also affects product design. Drafts need labels. Review packets need sources. Missing evidence needs to stay visible. Sensitive material needs matter-aware permissions. ABA Model Rule 1.6 on confidentiality is a useful reminder that information boundaries are part of the professional system, not an optional security feature.

Pilot Shape

A sensible pilot would start with two active matters and one review cadence.

Week one would map the current review process: where notes live, who prepares the meeting, which sources are authoritative, which deadlines matter, and what the partner wants to see. Week two would connect a narrow set of sources and produce the first review packet. Weeks three and four would refine categories, missing-context flags, and review queues based on how the team actually uses the packet.

The pilot should not measure whether AI can "do legal work." It should measure whether the team enters review with better context.

Useful success signals include:

• Less time spent assembling review notes.
• Fewer missed follow-ups.
• Clearer source links for review points.
• Better visibility of missing instructions or evidence.
• More consistent matter handover between lawyers.
• Partners reporting that review starts closer to the real issues.