DISCO

Dealing with a departed employee case? DISCO’s Senior Forensic Operations Specialist, Kim Reed, highlights how our digital forensics team identifies and secures data to protect your company's intellectual property. #discoprofessionalservices #digitalforensics #discoforensics #departedemployeecases #intellectualproperty

Raw AI models are basically a commodity now. Anyone with a credit card can access GPT or Claude. What does it take to build something lawyers can actually trust? First, it takes deterministic workflows. A general AI chatbot is fine for drafting a memo. But high-stakes litigation is different: legal work needs to be predictable and rule-based. Good legal tech wraps the AI in structured workflows so lawyers get consistent results every time. Then there's the need for private data. AI is useless without context. DISCO connects the engine to a firm's actual ediscovery documents. The AI works on that specific, secure data and nothing else. And finally, software alone doesn't solve complex enterprise problems. You can't just hand a firm a login and wish them luck. You need end-to-end professional services — consultants, project managers, support — to help run these tools. OpenAI and Anthropic build engines. DISCO builds the car.

Here are the numbers on our latest Auto Review. 📈 Precision was the name of the game for this project. Our team utilized DISCO's Auto Review to reach a 91% precision rate, significantly reducing the time spent on manual review. 📊 Key stats: 75,000 documents 📑  91% precision 🎯  96% recall 🔎 Answers to FAQs about our Auto Review process: - Our team typically takes 2 weeks to complete a project, regardless of size. (Time spent does not impact the metrics.) - We validate the results two ways: First, prior to running Auto Review, we take a control set from the entire corpus stratified by doc type. We use this set to validate we’re done with iteration and our metrics are good enough to run the full Auto Review. - Second, we take two sample sets after Auto Review completes, one from the positives (precision sample) and one from the negatives (elusion sample). Those allow us to calculate true positives and false negatives, which then allow us to calculate recall. Let us know if you have questions, and our team will be happy to respond! #autoreview

Very excited to be a part of ACEDS and look forward to leveraging my network to grow the North Carolina chapter. Especially given the current ediscovery landscape, reinventing itself with AI; it's a pivotal time to get together as a community to share ideas, best practices and lift each other up. Please reach out to me directly to join a group of like-minded legal technology professionals across the beautiful state of North Carolina.

Data-heavy. Compliance-driven. Expert-intensive. Three compound phrases describing one extremely complex arm of litigation. 💪 See how DISCO helps.

We're wrapping our coverage of our research into GenAI adoption in legal with a paradox. What do you do when 20% of legal professionals consider LACK of AI a dealbreaker... but 70% of law firms consider privacy a top concern? 👇

How does DISCO help your team handle complex data challenges? Brent Westenfelt, Senior Director of Data Operations & DISCO Desk, shares how our experts provide white-glove service to ensure your data is handled with precision. #withyouineverycase #discoprofessionalservices

We're continuing in our coverage of our new research into generative AI adoption in legal. Today: How are law firms actually using AI? The answer goes beyond review 👇

Another win for DISCO's Managed Review team! 🚀 This project achieved an astounding 99% recall rate in a GenAI-powered review of 244,000 documents. 📊 Key stats:  244k documents 📑  94.5% precision 🎯  99% recall 🔎 Answers to FAQs about our Auto Review process: - Our team typically takes 2 weeks to complete a project, regardless of size. (Time spent does not impact the metrics.) - We validate the results two ways: First, prior to running Auto Review, we take a control set from the entire corpus stratified by doc type. We use this set to validate we’re done with iteration and our metrics are good enough to run the full Auto Review. - Second, we take two sample sets after Auto Review completes, one from the positives (precision sample) and one from the negatives (elusion sample). Those allow us to calculate true positives and false negatives, which then allow us to calculate recall. Let us know if you have questions, and our team will be happy to respond! #autoreview

This week, we're deep-diving into our new research into generative AI adoption in legal. What's the real AI adoption story for firms and legal teams? Click to see 👇