Harvey and Claira in practice
By Kate Clark
In part two of this article, we look at setting up a workflow that carries through to production.
:quality(80)/f/285684285272562/1920x1080/ffa6ca44cb/2.png)
After 30 years’ experience leading and working with eDiscovery teams and technology, I believe the challenge in eDiscovery isn’t simply finding technology that can analyse documents.
It’s ensuring that the work is carried through the matter in a way that is consistent, efficient and defensible.
In part one of this article, I looked at how Claira, working within Nuix Discover, can extend the value of AI for firms already using Harvey.
Here I want to look at the practical side: how legal teams can set up a workflow that lets each tool do what it does best, from case preparation through to production.
Start with the output you need
For any task, I would start by asking, what does the team need at the end of it?
If the output is an analysis of correspondence, research or a draft based on selected evidence, Harvey supports that work within its workspace. If the output is a an issue based chronology, a reviewed document set with relevance, issue and privilege coding, and reasoning recorded against every document, Claira does that work in Nuix Discover.
If the output is a production for disclosure, it needs to be completed through Nuix Discover or RelativityOne. Analysis in Harvey can inform the team’s decisions, but the final production selection needs to be reflected in the review platform.
The handover is where Harvey and Claira connect
One of the opportunities I see is using the work already undertaken in Harvey to make a first-pass document review through Claira more effective.
Imagine a commercial dispute involving thousands of emails, contracts and internal documents. The legal team may use Harvey to analyse the pleadings, identify key parties, develop a chronology of events and clarify the issues that need investigation. That information can then help inform Claira’s Case Context and the review criteria applied to documents in Nuix Discover. Claira can use this context to assess documents against relevant issues and record coding decisions, and supporting reasoning directly in the review environment.
Once the review has identified documents relevant to particular issues, events or witnesses, the legal team can select those documents for further analysis, preparation or drafting in Harvey.
Any resulting decisions that affect coding, privilege, redactions or production selection can then be reflected in Nuix Discover, where the final production is managed. Rather than duplicating work, the two tools complement each other, with the outputs from one informing the next stage of the matter.
There is no automated integration between the platforms. This is a deliberate workflow that allows legal teams to use the strengths of each tool while maintaining control over the documents and decisions.
Agree on the roles at the start
Getting the most from both tools requires a clear understanding of where each task should happen, where its results need to be recorded and who is responsible for checking the outputs. For teams using Claira and Harvey together, I recommend agreeing on a few things at the start of the matter:
What is the required outcome? Are you developing an understanding of the issues, identifying relevant evidence, coding a document population or preparing documents for production?
Where should the work happen? Which tool is best suited to the task, and where do the results need to be recorded?
How will information move between the tools? What context, instructions and documents need to be shared, and how will that transfer be managed?
How will the data be protected? Security, confidentiality and data governance need to be considered whenever information moves between platforms. Claira does not store documents or scan answers, never uses them for model training and offers Australian data residency.
Who is responsible for checking the results? AI can accelerate the work, but lawyers and reviewers remain responsible for assessing outputs and making appropriate legal decisions.
How will the team avoid repeating work? Consistent terminology, clear instructions and well-managed handovers help ensure that work undertaken in one platform adds value to the next stage.
Getting these decisions right really can make a significant difference to how efficiently a matter progresses. It also helps teams avoid the situation where they have several powerful AI tools but no clear processes for using them.
The right people doing the right work
The same thinking applies to the people involved:
Lawyers bring the legal judgment by defining the issues, setting review criteria and making decisions on relevance and privilege.
Reviewers apply those instructions, check results and raise questions where the evidence needs closer attention.
Litigation support and eDiscovery specialists manage the review environment, keep coding consistent and prepare and validate the production in Nuix Discover.
Those roles may overlap but clear ownership matters. Getting value from either tool depends on having the right people directing the work, checking the outputs and taking responsibility for each stage.
Making Claira and Harvey work together
I don’t see Claira and Harvey as an either/or decision. Both have valuable capabilities and with clear roles, the right people and a deliberate handover, they can support a connected process from case preparation through to production.
At eDiscovery Collab, the exclusive provider of Claira in Australia and New Zealand, we are helping legal teams explore how Claira fits alongside the technology they already use.
If you’d like help mapping this workflow onto your next matter, I’d be happy to show you.
:quality(80)/f/285684285272562/756x667/c4c045a5f7/kate-clark.jpg)
Kate Clark
CEOKate is a senior eDiscovery expert with over 30 years of experience dedicated to simplifying complex legal data and streamlining operations for lawyers.
View full profile:quality(80)/f/285684285272562/1920x1080/6c2589faf7/1.png)
:quality(80)/f/285684285272562/3999x2490/559b3e668e/istock-1365567359-1.jpg)
:quality(80)/f/285684285272562/8192x5464/1f98ea0a0d/istock-1409451746.jpg)