Lexi

What do they actually do

Lexi builds AI software for law firms and in‑house legal teams that ingests emails, PDFs, medical records, transcripts and calls, extracts key facts and timelines, and generates first‑pass drafts for items like letters, motions, contracts and client updates. Current modules include intake (Screener), document analysis (Scan), drafting (Craft), lightweight case management (Track), and transcription/notes (Scribe). Outputs are designed to be attorney‑reviewed, not fully autonomous. The company says it’s used by “hundreds” of firms and, per YC, has processed 135,000+ documents across 7,000+ cases to date Lexi site YC company page.

For security and deployment, Lexi offers cloud and on‑prem options, says it does not train models on customer data, and uses multiple model providers (e.g., OpenAI, Anthropic, and open‑source) with contractual zero data retention; it is working toward SOC 2 and advertises 100+ integrations including Microsoft Word, Outlook, Google Docs and iManage Data Controls Lexi site YC company page.

Who are their target customer(s)

• Small-to-mid plaintiff firms handling heavy discovery and medical records: Manual intake, medical-record review and timeline building slow teams down and create errors; they want automated extraction and searchable case files so paralegals and attorneys spend less time on triage and summarizing.
• Transactional teams and contract reviewers at law firms: High volumes of routine contracts require quick clause detection, risk flags, and playbook‑consistent redlines to avoid repetitive, manual review work.
• In‑house legal teams at companies: They need fast, consistent answers for standard agreements and playbooked approvals while preserving privilege and strict data controls (often preferring on‑prem or zero‑retention providers).
• Large law firms / enterprise legal operations: Adoption is gated by tight DMS/editor integrations, SSO/SSO, audit trails, and SOC 2; lack of enterprise controls and permissioning blocks rollout.
• Solo and small‑firm attorneys trying to scale: Limited staff and billable pressure make first‑draft generation, intake screening, and automated task lists valuable so one lawyer can handle more matters without immediate hiring.

How would they acquire their first 10, 50, and 100 customers

• First 10: Run high‑touch pilots with plaintiff shops and boutique transactional teams recruited via YC/referrals and LinkedIn; ingest live matters, configure playbooks, and deliver a one‑page ROI summary (hours saved, cycle time).
• First 50: Package the pilot as a fixed‑scope “triage sprint” with onboarding and templated playbooks; use case studies plus targeted outbound (email/LinkedIn), CLE talks, and early‑customer referrals to convert similar firms.
• First 100: Add enterprise checkboxes (SOC 2, SSO, on‑prem, audit trails) and productize integrations/playbooks for common flows (NDAs, discovery triage); expand via DMS/editor partners, law‑firm tech consultants, and a referral/reseller program.

What is the rough total addressable market

Top-down context:

Per BLS, about 445,000 lawyers work in the U.S. legal services industry (law firms), with additional in‑house counsel tracked separately by ACC’s population tracker BLS OEWS ACC in‑house tracker.

Bottom-up calculation:

Focusing on U.S. law‑firm users only: 445k lawyers × 60% in small/mid‑market litigation/transactional roles Lexi targets × 50% near‑term serviceable × ~$1,800 ARR per seat ≈ ~$240M SAM (excluding in‑house).

Assumptions:

• 60% of law‑firm lawyers are in small/mid‑market litigation or transactional roles aligned to Lexi’s workflows.
• 50% of those are serviceable in the near term given compliance and change‑management constraints.
• Average pricing of ~$150/user/month (~$1,800 ARR) for core modules.

Who are some of their notable competitors

• CoCounsel (Thomson Reuters): AI assistant embedded in Westlaw/Practical Law for research, document analysis, and drafting. Strong on authoritative content; less focused on end‑to‑end intake + discovery workflows for plaintiff firms.
• Harvey.ai: Legal‑specific AI assistant with document review, research, and configurable workflows (Vault, Workflows). Positions toward enterprise rollouts and firm‑trained models/agents—overlaps with Lexi’s “AI associate” goal.
• Relativity (RelativityOne): Enterprise eDiscovery and review platform with AI for large‑scale processing. Overlaps on ingestion and extraction but is centered on full eDiscovery pipelines vs. lightweight intake/drafting.
• Logikcull: Cloud eDiscovery for small/mid firms with automated ingestion, culling, search, and A/V transcription. Competes on discovery speed and simplicity; narrower than Lexi’s multi‑module associate approach.
• Evisort (Workday Contract Intelligence): AI‑first CLM for clause extraction, obligations, templates, and enterprise workflows. Overlaps on playbooks/risk flags; focuses on end‑to‑end CLM rather than mixing litigation intake with contract review.