# **The Most User-Friendly AI Legal Tool Created by a Non-Legal Team**
[](https://mp.weixin.qq.com/mp/appmsgalbum?__biz=Mzg2OTY0MDk0NQ==&action=getalbum&album_id=4022587657917186053&subscene=126&scenenote=https%3A%2F%2Fmp.weixin.qq.com%2Fs%3F__biz%3DMzg2OTY0MDk0NQ%3D%3D%26mid%3D2247514246%26idx%3D1%26sn%3Dde104fc02cb120d9180f47b22fa4a1ef%26chksm%3Dcf71d6a76892668cf18f007c4c9974e8d4c946ed18281f0143b7aedede4e175cc1910467c9ba%26scene%3D126%26sessionid%3D1750994837%26subscene%3D227%26clicktime%3D1750994839%26enterid%3D1750994839%26key%3Ddaf9bdc5abc4e8d03659350f1d5aeabe879f96bc6ba36414330ae668002f202b47d03b5db81c8b59665b83c1f8583ed968cbf8713d565f7013dadcb18b5804616bb5ce7883c006475ceef002b63ded413b8449bda249ab3abfe984c0cd65c1607e949c660fbedd2861c0575826fd540c6b63d48daf52d266776cd0faaf14674e%26ascene%3D78%26uin%3DMzQ2MTE3Mjg4MA%253D%253D%26devicetype%3DiMac%2BMacBookAir10%252C1)
---
## Introduction
Legal work is one of the fastest-growing verticals for AI adoption. **Harvey** has surged to an $8B valuation by delivering deep, specialized products.
The Swedish newcomer **Legora** enables **lawyers and AI to work in parallel** in the same system — now used by tens of thousands of lawyers in leading global firms like *Linklaters* and *Cleary Gottlieb*.
- **Founded:** 2023, Stockholm
- **Series C Funding:** $150M (Oct 30, 2025) led by *Bessemer Venture Partners*
- **Valuation:** $1.8B USD
This article combines insights from founder and CEO **Max Junestrand**, interviews with YC Partner **Gustaf Alströmer**, the *Not Another CEO* podcast, and B2B user case studies.
---
## **01 — The Era of LLMs in Legal Tech**
In the LLM era, AI in legal tech is shifting from automation tools to **collaborative partners**.
**Key takeaways for devs without domain expertise:**
1. Embed domain expertise through iterative feedback.
2. Focus on user-centric design in vertical niches.
3. Avoid single-model dependency — leverage distribution and workflow integration.
### **Market Evolution**
- Pre-LLM legal tech: fragmented tools (contract generation, translation, redlining).
- Post-GPT era: semantic editing, unified model-based workflows, high-scale document ops.
- **Due diligence:** Reduced from hours to minutes; shifting from manual execution to **AI-managed reviews**.
### **Impact on Lawyers' Roles**
- Lawyers become **managers/reviewers** of AI outputs.
- Expertise remains essential for compliance & quality.
- Clients value rapid iteration and innovation over static solutions.
---
## **02 — Collaborative Lawyer AI Workspace**
### **About Legora**
- **Original Name:** Leya (rebranded Feb 2025)
- Early AI: Google BERT — failed in Swedish corpora, pivoted to GPT-3.5.
- Built *Stock Option Reader* then shifted focus: **End-to-end collaborative workspace**.
**Data compliance advantage:**
- Europe-only storage.
- No human review.
- Data excluded from model training.
**Client base:**
- 40+ markets, 400+ clients.
- Tens of thousands of lawyers.
- Funded $80M Series B (May 2025), valuation $675M.
- Offices: Stockholm, London, New York.
---
### **Core Features**
Legora offers **Web App**, **Word Plugin**, and **Playbook** — covering **review, drafting, research**.
#### **1. Web Application: Multi-Agent Collaboration**
- Started as document-centric chat; evolved into **intelligent agent system**.
- Executes multi-step workflows:
1. Research law.
2. Organize results in firm-standard language.
3. Output memorandum.
**Tech:**
- MCP architecture adapts to practice areas (IP, dispute resolution).
- **Tabular Review** runs 100K+ parallel queries with citation accuracy.
- Uses **chunking + RAG** for structured contract review.
#### **2. Word Plugin: AI Inside Native Tools**
- Embeds AI directly in MS Word.
- Features: template filling, anonymization, optimization, translation.
- Enables revision commands like:
`"Re-draft this MSA from the buyer's perspective"` referencing precedents.
#### **3. Playbook: Modular Legal Rules**
- Lawyers define rules to auto-evaluate clauses.
- Detects noncompliance, suggests fixes.
- Expands into sales, compliance, risk control.
---
## **Technology Foundation**
**Architecture:**
- Early alignment with Azure/OpenAI GPT.
- Now multi-cloud: AWS, Azure, Claude, Gemini, GPT, Mistral.
- **Hot-swap models** without upper-layer rebuild.
- Auto-inherits base model improvements.
**Efficiency:**
- Lightweight classifier routes requests to optimal model.
- Balances cost and response speed.
---
## **03 — Business Model & Sales**
### **Selling to AI Skeptics**
- Law firms fear reduced billables.
- AI reframed as **scale for expertise**, not a threat.
- Allows same team to serve far more clients.
- **Positioning:** Partner in transformation, not a replacement.
### **Sales Strategy**
- Buyers ≠ end-users; large firms use innovation depts.
- Proof-of-value pilots in due diligence, multi-doc review, research.
- Flexible *seat-based* licensing.
- Training-focused rollouts:
- BAHR: 4-month multi-training
- Bird & Bird: firm-wide in ~6–7 months
- Pérez-Llorca: 35+ trainings
---
## **04 — Competition & Risk**
### **Traditional Giants**
- Examples: Thomson Reuters, Littler, LexisNexis.
- Giants have strong content/distribution, but slow iteration cycles.
- Legora grows faster with small, adaptable team.
### **Harvey AI**
- Main modern competitor.
- Harvey got early GPT-4 access and funding.
- Legora focused on **fast product deployment** over costly fine-tuning.
- Both now default contenders in law firm AI procurement.
---
## **05 — Startup, Organization & Culture**
### **Founder Story**
Max Junestrand — competitive gamer turned entrepreneur.
Dual-enrolled in engineering & business schools.
Worked in esports analytics, startup accelerator Norrsken, McKinsey.
### **Team of Outsiders**
No legal background initially.
Interviewed hundreds of lawyers for workflow insights.
Early sales run personally by Max — grew ARR to $2M.
**Sales = learning in unknown markets.**
---
### **Hiring & Growth**
- Growth: 25 → 100 team members in 6 months.
- Offices: Stockholm, London, New York.
- Preference for **founder-type hires** with high growth trajectory.
- Flat org; AI-trained marketing team at 10× productivity.
**Interview filter:**
- Marketing: live product pitch.
- Engineering: build Legora POC + design explain.
---
## **Global Expansion**
- Started in Nordics for controlled scaling.
- Replicated into 15 markets before US entry.
- **Sequence:** Validate product-market fit → partnership → scale.
- Offices: US, Australia, planned India.
- **Localization:** Security, compliance — not language.
- Example: German cloud law exemptions with Microsoft, Google, AWS.
---
## **Advice for Vertical AI Founders**
1. **Avoid single-model lock-in.**
2. Focus on niche expert domains.
3. Embed creative model usage in unreplicable workflows.
4. Build systems to ride LLM evolution tides.
5. Team trust and collaboration > individual push.
---
---
## Extended Reading
[](https://mp.weixin.qq.com/s?__biz=Mzg2OTY0MDk0NQ==&mid=2247519279&idx=1&sn=38988fa076d39737e15c66ac03ef725b&scene=21#wechat_redirect)
---
## Related Discussions
[](https://mp.weixin.qq.com/s?__biz=Mzg2OTY0MDk0NQ==&mid=2247519240&idx=1&sn=30efcbe447c20697b48ae09291da43cf&scene=21#wechat_redirect)
---
[](https://mp.weixin.qq.com/s?__biz=Mzg2OTY0MDk0NQ==&mid=2247519112&idx=1&sn=562af823af24b984bc3c0242c389caa8&scene=21#wechat_redirect)
---
[](https://mp.weixin.qq.com/s?__biz=Mzg2OTY0MDk0NQ==&mid=2247519023&idx=1&sn=54a12b29668ca5b99582d8d4b931125f&scene=21#wechat_redirect)
---
[](https://mp.weixin.qq.com/s?__biz=Mzg2OTY0MDk0NQ==&mid=2247518893&idx=1&sn=d97353a4eaff11bf3a12a3712ca328db&scene=21#wechat_redirect)
---
[Read Original Article](2247519321)
[Open in WeChat](https://wechat2rss.bestblogs.dev/link-proxy/?k=dfe27ba5&r=1&u=https%3A%2F%2Fmp.weixin.qq.com%2Fs%3F__biz%3DMzg2OTY0MDk0NQ%3D%3D%26mid%3D2247519321%26idx%3D1%26sn%3Dfc53c50e23296ab6d3504468e80f0785)
---
