Case Study: Transforming Legal Investigation with AI-Powered Insight
-min.avif)
Fennemore Craig, one of the nation’s fastest growing law firms known for litigation and investigations, recently completed a document review project involving 15,000 documents. While the team successfully identified key findings, the review took 10 days, consuming significant attorney resources. As a leader in adopting emerging AItools, the firm turned to TrueLaw, now a Consilio company, to test a new approach.
In a matter of minutes, AI Narratives replicated what had previously taken a ten-day manual review.
Challenge: The Limits of Prompt-Based Investigation
Like many teams handling increasingly complex datasets, Fennemore experimented with large language models (LLMs) and prompt engineering to streamline early-stage investigations—but surfacing meaningful insight, quickly and consistently, remained elusive. Traditional prompt engineering with LLMs has emerged as a tempting solution, but in practice, it’s proven inefficient, unpredictable, and costly. Teams can spend days manually crafting and refining prompt iterations, hoping to land the right one that surfaces something meaningful. And even when something useful surfaces, critical questions remain unanswered.
What about the insights not previously considered? What about the blind spots buried in the dataset that manual prompting may not anticipate? This approach can be slow, reactive, and misaligned to the speed, flexibility, and complexity of today’s modern eDiscovery matters.
There had to be a more strategic, scalable way to harness the transformative power of LLMs, and that was the hypothesis behind TrueLaw’s AI Narratives.
Solution: AI Narratives Reframes the Investigation
To test the potential of this new tool, Fennemore’s eDiscovery team used AI Narratives on a15,000-document corpus from a previously closed matter. Their goal was to determine whether TrueLaw could identify the most critical insights from a native, unlabeled dataset using just a high-level prompt as compared to the traditional manual review method. It was a real-world, ground-truth scenario. They already knew what was in the data, and how long it had taken to find it.
TrueLaw developed AINarratives to eliminate the guesswork and supercharge the early analysis of large, complex datasets. Rather than relying on prompt after prompt, AINarratives enable the user to input a single, high-level query. From there, a network of automated reasoning systems generates follow-up queries, interrogates the pre-indexed (vectorized) dataset, and builds a dynamic picture of what’s actually in the data—surfacing not just what you asked for, but what you didn’t know to ask. This process reframes the investigation intelligently, using AI’s deep analysis capabilities to surface the most meaningful content for legal analysis.
With AI Narratives, it took just a few hours to index the data and 30 minutes to generate actionable insights. The system delivered an interactive, report-ready output detailing relevant custodians, hot documents, thematic timelines, and suggested paths for further investigation. Every insight was anchored to its source material, fully linked, referenceable, and transparent, giving attorneys the confidence to rely on AI without fear of hallucinations.
The tool brings the lawyer’s expertise into harmony with AI’s scale and speed. It also eliminates the trial-and-error loop of trying to master prompt engineering. Because AINarratives integrates directly into leading review platforms like RelativityOne and Relativity Server, findings can be immediately ported back into there view platform, enabling a hyper-focused and ultra-efficient document review workflow.
Outcome: Precision, Speed, and Unexpected Discovery
The tool brings the lawyer’s expertise into harmony with AI’s scale and speed. It also eliminates the trial-and-error loop of trying to master prompt engineering. Because AINarratives integrates directly into leading review platforms like RelativityOne and Relativity Server, findings can be immediately ported back into the review platform, enabling a hyper-focused and ultra-efficient document review workflow.
The results exceeded expectations. Not only did AI Narratives match the human team’s original findings, it did so in considerably less time.
Fennemore described the experience as transformative:
“AI Narratives gave us a level of insight we usually wouldn’t see until much later in the matter. It identified the right documents, surfaced the right custodians, and even uncovered a key piece of evidence that was not considered or surfaced during the traditional review process. That kind of precision, delivered up front, has fundamentally changed how we think about data analysis. We’re now re-evaluating how we structure our workflows to take advantage of this kind of AI-driven acceleration.”
- Adrian D'Amico, Director of Emerging Technology & Innovation, Fennemore
Conclusion: From Discovery to Direction
With AI Narratives, Fennemore demonstrated what’s possible when legal expertise is amplified, not replaced, by AI. The result was not only faster and more complete, but more strategic: a process that empowered their attorneys to move quickly, act with confidence, and surface deeper insight than traditional approaches could deliver.
For legal teams facing growing data complexity and shrinking timelines, this wasn’t just a proof of concept. It was a glimpse into a smarter, more defensible way forward.
Fennemore Craig, one of the nation’s fastest growing law firms known for litigation and investigations, recently completed a document review project involving 15,000 documents. While the team successfully identified key findings, the review took 10 days, consuming significant attorney resources. As a leader in adopting emerging AItools, the firm turned to TrueLaw, now a Consilio company, to test a new approach.
In a matter of minutes, AI Narratives replicated what had previously taken a ten-day manual review.
Challenge: The Limits of Prompt-Based Investigation
Like many teams handling increasingly complex datasets, Fennemore experimented with large language models (LLMs) and prompt engineering to streamline early-stage investigations—but surfacing meaningful insight, quickly and consistently, remained elusive. Traditional prompt engineering with LLMs has emerged as a tempting solution, but in practice, it’s proven inefficient, unpredictable, and costly. Teams can spend days manually crafting and refining prompt iterations, hoping to land the right one that surfaces something meaningful. And even when something useful surfaces, critical questions remain unanswered.
What about the insights not previously considered? What about the blind spots buried in the dataset that manual prompting may not anticipate? This approach can be slow, reactive, and misaligned to the speed, flexibility, and complexity of today’s modern eDiscovery matters.
There had to be a more strategic, scalable way to harness the transformative power of LLMs, and that was the hypothesis behind TrueLaw’s AI Narratives.
Solution: AI Narratives Reframes the Investigation
To test the potential of this new tool, Fennemore’s eDiscovery team used AI Narratives on a15,000-document corpus from a previously closed matter. Their goal was to determine whether TrueLaw could identify the most critical insights from a native, unlabeled dataset using just a high-level prompt as compared to the traditional manual review method. It was a real-world, ground-truth scenario. They already knew what was in the data, and how long it had taken to find it.
TrueLaw developed AINarratives to eliminate the guesswork and supercharge the early analysis of large, complex datasets. Rather than relying on prompt after prompt, AINarratives enable the user to input a single, high-level query. From there, a network of automated reasoning systems generates follow-up queries, interrogates the pre-indexed (vectorized) dataset, and builds a dynamic picture of what’s actually in the data—surfacing not just what you asked for, but what you didn’t know to ask. This process reframes the investigation intelligently, using AI’s deep analysis capabilities to surface the most meaningful content for legal analysis.
With AI Narratives, it took just a few hours to index the data and 30 minutes to generate actionable insights. The system delivered an interactive, report-ready output detailing relevant custodians, hot documents, thematic timelines, and suggested paths for further investigation. Every insight was anchored to its source material, fully linked, referenceable, and transparent, giving attorneys the confidence to rely on AI without fear of hallucinations.
The tool brings the lawyer’s expertise into harmony with AI’s scale and speed. It also eliminates the trial-and-error loop of trying to master prompt engineering. Because AINarratives integrates directly into leading review platforms like RelativityOne and Relativity Server, findings can be immediately ported back into there view platform, enabling a hyper-focused and ultra-efficient document review workflow.
Outcome: Precision, Speed, and Unexpected Discovery
The tool brings the lawyer’s expertise into harmony with AI’s scale and speed. It also eliminates the trial-and-error loop of trying to master prompt engineering. Because AINarratives integrates directly into leading review platforms like RelativityOne and Relativity Server, findings can be immediately ported back into the review platform, enabling a hyper-focused and ultra-efficient document review workflow.
The results exceeded expectations. Not only did AI Narratives match the human team’s original findings, it did so in considerably less time.
Fennemore described the experience as transformative:
“AI Narratives gave us a level of insight we usually wouldn’t see until much later in the matter. It identified the right documents, surfaced the right custodians, and even uncovered a key piece of evidence that was not considered or surfaced during the traditional review process. That kind of precision, delivered up front, has fundamentally changed how we think about data analysis. We’re now re-evaluating how we structure our workflows to take advantage of this kind of AI-driven acceleration.”
- Adrian D'Amico, Director of Emerging Technology & Innovation, Fennemore
Conclusion: From Discovery to Direction
With AI Narratives, Fennemore demonstrated what’s possible when legal expertise is amplified, not replaced, by AI. The result was not only faster and more complete, but more strategic: a process that empowered their attorneys to move quickly, act with confidence, and surface deeper insight than traditional approaches could deliver.
For legal teams facing growing data complexity and shrinking timelines, this wasn’t just a proof of concept. It was a glimpse into a smarter, more defensible way forward.
.png)
.png)
