Legal document review · Production

Global technology leader achieves 75% cost savings and 88% faster document review with OpenText eDiscovery Aviator

The problem

A global technology company's legal team faced escalating eDiscovery costs, with document review consuming 75% or more of total eDiscovery expenses. Human reviewers required extensive training, supervision, and quality control, demanding significant time from senior attorneys and creating a bottleneck that prevented focus on strategic legal work.

First attempt

Traditional document review workflows were inefficient, requiring extensive reviewer training, continuous supervision, and iterative quality control, with project timelines often spanning months.

Workflow diagram · grounded in source
1
Review memo submitted as prompt
trigger
“The system required only the existing review memo—the same document already prepared for human reviewers—as its prompt. This approach bypassed weeks of reviewer training and ongoing quality control supervision”
2
LLM classifies document responsiveness
ai_action
“Gen-AI powered document review using a secure, pre-trained large language model to automate responsiveness review—minimizing human intervention”
3
Validation against benchmark
validation
“rigorous validation analyzing roughly 244,000 documents from a previously completed antitrust matter. Employing a comprehensive testing approach ensured the GenAI workflow could deliver reliable, defensible results that would meet judici…”
4
Defensible results produced
output
“deliver reliable, defensible results that would meet judicial standards”
Reported outcome

The AI-powered document review pilot achieved 75% cost savings and 88% faster completion time, compressing the review timeline from three months to a few days, while delivering classification quality that met or exceeded human-review benchmarks.

Reported metrics
Cost savings75%
Document review completion time improvement88%
Review timeline compressedfrom three months to just a few days
document review share of eDiscovery expenses75% or more of its total eDiscovery expenses
Show all 5 reported metrics
cost savings75%
document review completion time improvement88%
review timeline compressedfrom three months to just a few days
document review share of eDiscovery expenses75% or more of its total eDiscovery expenses
documents analyzed in pilotroughly 244,000
Reported stack
OpenText eDiscovery Aviator reviewlarge language modelAWS Bedrock
Source
https://www.opentext.com/customers/global-technology-leader
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

The AI-powered document review pilot achieved 75% cost savings and 88% faster completion time, compressing the review timeline from three months to a few days, while delivering classification quality that met or excee…

What tools did this team use?

OpenText eDiscovery Aviator review, large language model, AWS Bedrock.

What results were reported?

Cost savings: 75%; Document review completion time improvement: 88%; Review timeline compressed: from three months to just a few days; document review share of eDiscovery expenses: 75% or more of its total eDiscovery expenses (source-reported, not independently verified).

What failed first in this deployment?

Traditional document review workflows were inefficient, requiring extensive reviewer training, continuous supervision, and iterative quality control, with project timelines often spanning months.

How is this legal document review AI workflow structured?

Review memo submitted as prompt → LLM classifies document responsiveness → Validation against benchmark → Defensible results produced.

Related legal document review cases