DocLens.ai

Thank you, Arnab Dey and team, for sharing your journey and insights into building an enterprise-grade RAG product. Thanks as well for your generosity in providing access to DocLens, which gives our students invaluable hands-on experience in designing and conducting evaluations to assess RAG systems. Alex and I look forward to continuing our collaboration and exploring new opportunities together. S. Alex Yang and I thoroughly enjoyed working with you on this and look forward to continuing our collaboration.

Some news we are proud to share 🎉 DocLens.ai is now the subject of a case study at London Business School, written by Professors Nitish Jain and S. Alex Yang The case explores a question every enterprise AI company eventually faces. Getting clients to believe in what generative AI can do is only the first step. The harder one is earning trust in how reliably it does it, across high-stakes, document-heavy work like insurance claims. That shift, from promise to proven reliability, has shaped how we build. Context, evaluation, and customization are not features we added. They are the foundation. The internal metaphor that guided us from day one was a "Watson" for claims adjusters. A system that surfaces what matters, frames the context, and supports human judgment without ever overstepping it. To have that journey studied and taught at one of the world's leading business schools is a genuine honor. Thank you to the LBS team, and to the clients and colleagues who shaped this story with us. Read more: https://lnkd.in/gEudUhfq Arnab Dey, Michael Bruton, Amit Saha

Insurance defense litigation runs on document volume. A single case can arrive with thousands of pages of records, depositions, and expert reports, and weeks disappear into manual review before strategy even begins. That math is changing. We looked at where AI-powered file review actually pays off for defense firms, and four use cases stood out: medical record summaries, litigation timelines, deposition analysis, and expert witness review. Each one takes work off the desk where human judgment adds the least, and puts attorney time back on the work that wins cases. One principle holds it together: AI augments judgment, it does not replace it. Every output is verified by a legal professional. Read the full breakdown here: https://lnkd.in/dHVMZ3wW Arnab Dey Amit Saha Michael Bruton

Transform Complex Claims. Scout InsurTech 2026 brings together the people shaping what's next for insurance. At DocLens.ai, we are focused on a simple question: How can our platform help insurers navigate unstructured documents to identify complex-claims-that-matter-early while freeing up time for the claim professional to focus on the human side of claim processing? Arnab Dey, our Co-founder & CEO, will be at Scout InsurTech 2026 discussing the future of claims intelligence, medical record analysis, and AI-driven workflows. If you're attending, we'd love to connect. Michael Bruton, Amit Saha

Transform Complex Claims. Scout InsurTech 2026 brings together the people shaping what's next for insurance. At DocLens.ai, we are focused on a simple question: How can our platform help insurers navigate unstructured documents to identify complex-claims-that-matter-early while freeing up time for the claim professional to focus on the human side of claim processing? Arnab Dey, our Co-founder & CEO, will be at Scout InsurTech 2026 discussing the future of claims intelligence, medical record analysis, and AI-driven workflows. If you're attending, we'd love to connect. Michael Bruton, Amit Saha

Medical records tell a story, but finding the critical moments often means digging through hundreds or thousands of pages. AI-powered medical chronologies can change that by automatically organizing records into clear, structured timelines, helping teams move from document review to decision-making faster. Our latest blog explores how automated timeline creation can reduce manual effort, surface key events, and bring consistency to complex medical record analysis. Read more: https://lnkd.in/gBCdcctH Arnab Dey Amit Saha Michael Bruton

In liability claims, the truth is rarely found in a single document. It lives in the gaps between medical records, the evolution of claimant narratives, the timing of treatment and testimony, and the patterns repeated across claims. Traditional systems — and even knowledge graphs — can tell you *what* entities are related. They can't tell you *what it means* for liability, damages, or exposure. Context graphs change that. They model how evidence behaves over time within a claim, not just map it. In practice: → Auto liability: Chiropractic care starting 3 weeks post-accident, matching patterns from prior claims → General liability: A slip-and-fall claimant linked to similar claims at 3 other locations → Workers' comp: Physical therapy exceeding clinical guidelines by 300% → Med mal: Retrospective documentation changes after an adverse outcome The industry is moving from "What information do we have?" to "What does this information mean?" That's where better claims outcomes begin. Read the full blog here: https://lnkd.in/gQD-2xE3 #ClaimsManagement #AIEngineering #BuildingWithAI Arnab Dey, Michael Bruton, Amit Saha

We’re heading to the Medical Professional Liability Association Conference 2026 in Philadelphia. Join us in the exhibit area to see how DocLens.ai accelerates the review of claim documents for MPL and enables timely, accurate risk capture. Arnab Dey, Michael Bruton, Amit Saha

AI models are getting better. But better at what? We re-ran our liability claims benchmark with Claude Sonnet 4.6 to find out. The results are directionally interesting. Stronger performance on simple retrieval tasks. But when it comes to complex claim reasoning and extracting key financial details, gaps remain. In insurance claims, 60% accuracy is not good enough. It only improves with a stronger domain context layer and access to external data. That is exactly what ClaimLens™️ is built for. This carousel breaks down what has changed and what has not. Arnab Dey, Michael Bruton, Amit Saha