How Claira and CoCounsel Work Together

Your firm may already own CoCounsel. Thomson Reuters has built it into a broad assistant that now spans legal research, drafting, and document analysis, and the litigators we speak with often have it open in another tab while they are asking us about document review. The question that follows is a fair one. If CoCounsel can review documents, why add Claira?

The answer is not that one tool is better. The two products solve different problems, and they solve them in different places. Knowing where one stops and the other starts makes both of them more useful.

What CoCounsel Does Well

CoCounsel Legal brings research, drafting, and document analysis into a single assistant. Its Deep Research capability is grounded in Westlaw and Practical Law, which means the research it returns is anchored to editorially maintained authority rather than to whatever the model absorbed during training. Deep Research Verify goes a step further and checks whether the cited authority actually supports the assertion made, flagging misattributions before they reach a brief.

The skills library is where most litigators spend their time. Review Documents handles deposition transcripts, medical records, and contracts, and it answers questions that require inference rather than keyword matching. Summarize, Compare Documents, Timeline, and Tabular Analysis cover the analytical work that sits between reading a file and writing about it. For transactional teams, the agentic contract review layer takes on orientation and first pass markup, then hands the issues back for lawyer judgment.

This is good software, and none of what follows is a criticism of it. CoCounsel is built for a lawyer working with a curated set of materials in a research and drafting context. Within that shape, it performs.

Where the Handoff Has to Happen

The boundary is not intelligence. It is volume and location.

The Review Documents skill accepts two hundred files at one time. Thomson Reuters publishes that limit plainly, and it is the right limit for the work the skill was designed around. A due diligence set, a batch of transcripts, the twenty agreements that matter in a contract dispute: those are document sets a lawyer curates by hand before asking a question of them.

A discovery population is not curated by hand. After culling, a fifteen custodian collection can still leave two hundred thousand documents that each need a defensible decision recorded against them. Uploading that population two hundred files at a time is not a workflow. It is a data entry project.

There is a second limitation worth naming, because it matters more in litigation than anywhere else. Review Documents cannot identify Bates or line numbers directly. For a research task that is irrelevant. For a review that has to survive a meet and confer, a privilege log, and eventually a production, the identifier attached to each decision is close to the entire point.

Those documents already live in Nuix Discover, with family relationships intact, the coding panels your reviewers work in every day, and a production workflow everything has to pass through. Moving them somewhere else to be read is not a change of workflow. It is a migration, and migrations of evidence raise custody and completeness questions you will be asked to answer later. The more sensible move is to leave the population where it is and bring the review to it. We made the same argument about Harvey earlier this month, and the shape of it does not change from one assistant to the next.

Feeding CoCounsel's Thinking Into Claira

By the time a matter reaches first pass review, your team has usually done real thinking in CoCounsel. That thinking might live in a research memo grounded in Westlaw, a timeline built from the pleadings, or a tabular analysis that mapped the key agreements and the people who signed them.

That work product is exactly what Claira needs in order to review well. Claira's Case Context feature lets you give the AI background on the matter before any scan runs. It is structured into five sections: parties and people, description and timeline, relevance and issues, privilege indicators, and collection details. A case theory developed in CoCounsel translates almost directly into those fields. The names, dates, and project code words that surfaced in your research become the anchors that keep a bulk review consistent across every document in the population.

The same handoff works for review criteria. Ask CoCounsel to summarize the allegations, the defences, and the categories of documents each side will care about. Then distil that summary into the single substantive question a reviewer would ask of each document. That question, refined with exclusions for the routine traffic you do not want flagged, becomes your Claira prompt. We have published a full walkthrough of that translation, including the three artifacts worth lifting out of CoCounsel before you start.

Lawyers do the legal thinking where legal thinking is best supported. Claira applies it at scale, one document at a time, inside the platform that already holds the evidence.

Sending Claira's Output Back Up

The pattern is a loop rather than a pipeline.

Once a bulk scan finishes you have a coded population and, more usefully, a much smaller set of documents that actually matter. Two hundred thousand documents becomes a few hundred hot documents with reasoning recorded against each one. That set fits comfortably inside CoCounsel's design. It is curated, it is small, and it is close to the ideal input for the Review Documents and Timeline skills.

Export the hot set, load it into CoCounsel, and ask it to build the deposition outline, draft the chronology section of the brief, or test your strongest documents against the authority Deep Research surfaces. The narrowing happened where the volume lived. The drafting happens where the drafting tools live.

Where Each Tool Earns Its Keep

If you are budgeting for both, the split is straightforward. CoCounsel earns its keep on research, drafting, and analysis of curated materials, and on the Westlaw grounding that no eDiscovery platform provides. Claira earns its keep on the part nobody else wants to price: reading every document in a discovery population and recording a defensible decision against each one, without the data ever leaving Nuix Discover.

Neither replaces the other. Firms that run both tend to describe the same rhythm. Think in CoCounsel, review in Claira, then come back to CoCounsel to write.

If you already have CoCounsel and want to see what the second half of that loop looks like against your own document set, book a short session with our team. Bring a matter you know well. It is a faster conversation than a demo built on sample data.