Jay Madheshwaran

AI Summary

→ WHAT IT COVERS a16z partner Alex Rampell interviews Eve founder Jay Madheshwaran about how AI is transforming plaintiff law — the contingency-based legal sector serving everyday people. Eve's platform handles client intake through settlement, enabling attorneys to manage five times more cases without adding headcount, unlocking a historically underserved legal market. → KEY INSIGHTS - **Labor-to-revenue conversion:** Plaintiff law firms reject 96% of labor and employment cases because pre-case research alone takes 10+ hours per matter, with no guaranteed $5,000 minimum fee recovery. Eve compresses that intake and qualification work to near-zero time, making previously uneconomical cases viable and converting rejected leads into billable settlements. - **5x caseload multiplier:** One Los Angeles firm scaled from 100 clients per attorney to 500 by deploying Eve across the full case lifecycle — intake qualification, inter-team handoffs, document generation, and pre-mediation settlement briefs. Firms targeting this outcome should implement AI at every handoff point, not just intake, to capture the full capacity gain. - **Data moat in non-public settlements:** Less than 1% of plaintiff cases produce published legal opinions; the rest settle privately, leaving no scrapable dataset. Eve builds proprietary outcome data from each resolved case, creating jurisdiction-specific and opposing-counsel-specific intelligence that raw AI models like Claude or GPT cannot replicate without this embedded workflow context. - **Adversarial agent dynamics already emerging:** Defense firms use timing tactics — filing discovery requests the day before major holidays — to overwhelm plaintiff attorneys. One Eve customer built a dedicated agent to respond to discovery the same day regardless of timing, neutralizing the tactic entirely. Firms should identify their three most common opposing-counsel tactics and automate counter-responses first. - **Software market is a shadow labor market:** Plaintiff firms historically ran on a cell phone and email — no meaningful software spend. Because their revenue is outcome-based (typically one-third of settlement), and labor consumes roughly half of that share, AI that eliminates labor cost converts directly into margin. The addressable software market is therefore sized against hundreds of billions in annual settlements, not prior software budgets. → NOTABLE MOMENT Madheshwaran reveals that a single personal injury case involves roughly 10,000 discrete actions — outbound calls to insurers for medical records, document exchanges, negotiation rounds — and that Eve's voice AI agents are already encountering and learning to navigate other AI agents on the defense side during routine calls. 💼 SPONSORS None detected 🏷️ Legal AI, Plaintiff Law, Agentic Workflows, Access to Justice, AI Business Models