OpenText eDiscovery Aviator Flex: Turn prompts into lasting legal intelligence

GenAI interest is rising. Usage of GenAI in the legal industry nearly doubled from 14% in 2024 to 26% in 2025[1]. Legal teams understand that they do not need GenAI to replace legal judgment, but to help apply that judgment more efficiently, consistently, and confidently across defined document sets.
 
Attorneys often already know the questions they need answered: what happened, who knew what and when, which documents support or contradict a position, and where the relevant obligations, risks, or admissions appear. The challenge is not identifying those legal questions but translating them into instructions that consistently produce useful results across a matter.
 
This is especially true because, up until now, legal teams have been using GenAI to perform prescribed tasks. What if you were no longer limited by the tasks mapped out for you, but could ask any question of your data that you need answered? What if you could query as much or as little of your data set about whatever you needed to know, whenever you needed to know it?
 
Let’s take a look at the challenges and how OpenText eDiscovery Aviator Flex addresses them.

Overcome fears around prompt engineering

The growing focus on prompt engineering to support eDiscovery solutions can make GenAI seem like a separate technical discipline that attorneys must master before they can use it effectively.  But that is not true. Attorneys do not need to become prompt engineers. What they need is a more accessible, repeatable way to translate the questions they already know how to ask into effective GenAI-assisted analysis – a way to ask questions in natural language and conduct “on-the-fly” exploration without a predefined workflow.

Legal teams already examine documents to identify what matters. They extract facts, trace events, identify key people, evaluate issues, test case theories, locate contractual obligations, assess risk, and build chronologies. Every review challenge requires additional extraction of information, insights, facts, events, people, or issues from documents.

That work may currently happen through manual document review, keyword searches, technology-assisted review, or a combination of approaches. The task is not new. What changes is the ability to apply a question across a defined document set and capture the results in a more structured form and drive consistency across the process. The answer is not simply to give every attorney a blank prompt box and expect consistent results. One attorney may provide detailed context and request a clearly structured response. Another may ask a shorter version of the same question and receive a different output. A third may develop an effective prompt but leave it buried in personal notes or a one-time interaction.

That is experimentation. It is not yet a repeatable legal workflow.

Use Aviator Flex to run pre-built or custom prompts

The recently launched OpenText eDiscovery Aviator Flex feature enables users to run a custom AI prompt across any document set and save the results directly to a custom text field in the database. Unlike predefined Aviator features, Aviator Flex gives legal teams the freedom to ask exactly what a matter requires, while making AI-generated insights persistent, searchable, and available across the team for review, reporting, and trial preparation.

Aviator Flex is intended for controlled, repeatable enrichment tasks such as document summaries, issue statements, or structured extractions. Attorneys do not have to master a growing catalogue of prompt-engineering techniques simply to ask whether documents reveal:

• when a custodian became aware of an event
• whether communications reference a particular project or agreement
• which documents contain admissions or contradictions
• where contractual termination or indemnification language appears
• which files identify personal information or potential privilege; or
• what events should be added to a chronology.

You no longer need to assemble a whole arsenal of custom tools for “niche” use cases.  One single tool can provide endless ways to structure your insight. That flexibility allows the same underlying capability to support different forms of legal analysis, including:

• early case assessment
• customized responsiveness and issue coding
• case strategy formulation
• contract analysis
• investigations
• data subject access requests
• data breach investigations

The point is not that every matter requires all of these uses. It is that legal teams can adapt the inquiry to the needs of the matter instead of being limited to a pre-determined workflow offered by its eDiscovery solution.

Aviator Flex provides reusable intelligence

A legal team may spend time refining a question so it identifies a particular category of risk, extracts events in a consistent format, or distinguishes between documents that merely mention a contract and those that discuss an obligation under it. Once that prompt works, the team should not have to recreate it every time.

Prompting becomes more valuable when it stops being disposable.  A useful AI answer is only the beginning. If the response remains in a chat window, gets copied into an email, or disappears into an attorney’s notes, the organization loses much of its value. The insight cannot easily be searched, compared, shared, filtered, or incorporated into the broader review.

Aviator Flex not only allows legal teams to ask questions in natural language, design and run custom prompts against selected documents, but also save prompts for reuse within the same project and store output directly to a custom or existing text field for use in a broader review.  Flex helps you capture that intelligence in a structured, consistent way. You get a jump on traditional ad hoc review notes and can “red flag” key items on day one.

With Aviator Flex, outputs can be stored in custom fields associated with the documents being analyzed. A field called “Custodian Awareness” might summarize what each document indicates about what a person knew and when. A “Contract Reference” field might record whether the document discusses a particular agreement and explain the connection. A “Risk Indicator” field might identify language suggesting potential legal, regulatory, financial, or reputational exposure.

Once captured in a field, the output becomes structured and persistent. Reviewers can see it alongside the document. Teams can search, sort, filter, report on, and export the information through established workflows. Insight generated can remain available during later review, deposition preparation, and case strategy. This is what separates a one-time GenAI response from reusable matter intelligence.

From experimentation to reusable legal intelligence

Aviator Flex does not ask legal teams to reinvent how they work. It takes the analysis they already perform – extracting facts, events, people, issues, risk, and meaning from documents – and helps make it structured, repeatable, and reusable through prompts.

That is the power of Aviator Flex. This is Aviator functionality without boundaries — you get full flexibility to choose the scope of the data queried and the freedom to choose the question you want Aviator to answer about your data. 

One GenAI tool, endless ways to structure your insight. Write the prompt. Define the document set. Aviator does the rest.

See how OpenText eDiscovery Aviator can help you use GenAI to transform case strategy, document review, and investigations for enterprise legal teams and law firms alike.

[1] Thomson Reuters, 2025 Generative AI in Professional Services Report, 2025.