Max Junestrand on Building Legora and Learning Faster

The Legora co-founder discusses customer research, rebuilding the product, team culture and evaluating AI for legal work.

MJMax Junestrand登壇者 · Legora
チャンネル
Y Combinator
公開日
2026年8月25日
動画
59:39
読了時間
5 分
原文
英語
YouTube で開く
スクロールすると動画は右下に小さくなって再生を続けます。字幕は閲覧言語に合わせて切り替わります。
閲覧言語2 言語で読めます
この記事はまだ日本語に翻訳されていないため、英語の原文を表示しています。
要約

30 秒でわかる

At YC Startup School, Legora CEO Max Junestrand describes the company’s early experience, a sales pause to rebuild its product and lessons on people and growth. Usage and revenue figures are presented as company-reported milestones.

  1. 1Pay attention to customersThe founders interviewed lawyers and observed their work.
  2. 2Rebuild when necessaryThe company paused sales to address product weaknesses.
  3. 3Learn the domainThe founders learned from users; legal work still needs professional expertise.
  4. 4Evaluate the real taskLegal evaluations need to reflect workflows and consequences of errors.
  5. 5Treat growth figures as reportedThe revenue milestone is a company-reported figure.

ポイント

  1. Direct customer research helps founders understand real workflows.
  2. Rebuilding a weak product may require difficult tradeoffs.
  3. Hiring and culture depend on more than credentials.
  4. Legal AI needs task-specific evaluation and professional review.
  5. Growth figures should be read as time-specific company reports.
全文

チャプターごとに整えた全文を読む

文字起こしのみ
01

Finding the customer problem

Legora co-founder and CEO Max Junestrand speaks at YC Startup School about building AI software for lawyers. He shares the company’s early customer research, product decisions and lessons about teams.

Legora co-founder and CEO Max Junestrand speaks at YC Startup School about building AI software for lawyers. He shares the company’s early customer research, product decisions and lessons about teams.

Junestrand describes legal work as an opportunity for software because many professionals spend time searching, summarizing and drafting. His company’s usage and revenue figures are company-reported, time-specific measures.

He shares customer feedback to illustrate the value the team hoped to provide. Individual anecdotes can reveal user needs, but they do not independently establish accuracy or suitability for every legal matter.

02

Learning from lawyers

The idea began when his co-founder saw law students summarizing court cases during internships. The team wondered whether language models could reduce that repetitive work.

The idea began when his co-founder saw law students summarizing court cases during internships. The team wondered whether language models could reduce that repetitive work.

Junestrand joined the project after seeing an early demonstration. The founders had software experience but needed to learn how lawyers actually worked and what counted as a useful result.

They contacted lawyers and offered to pay for their time to learn from them. Early conversations were research, not just sales: they helped the team identify language, workflows and pain points.

The team spent time in a law firm’s office to observe work closely. Being near users gave them feedback that would have been hard to obtain from assumptions alone.

03

Rejection and fundraising

The founders initially received a YC rejection. Junestrand recalls questions about their understanding of the market, and says the experience exposed gaps they needed to address.

The founders initially received a YC rejection. Junestrand recalls questions about their understanding of the market, and says the experience exposed gaps they needed to address.

Continued customer work gave them more knowledge than they initially realized. Over time they learned to connect a technical demonstration to the practical needs of lawyers.

Junestrand describes moving to San Francisco and meeting potential investors. He says the experience forced him to explain the company clearly and keep refining its story.

Storytelling, in his view, is not merely a fundraising performance. A founder must communicate why the problem matters and recruit people who want to work on it.

04

Pausing sales to rebuild

The company later paused sales for several months to rebuild the product. He says early growth had revealed weaknesses in the experience that the team needed to address.

The company later paused sales for several months to rebuild the product. He says early growth had revealed weaknesses in the experience that the team needed to address.

A pause carries a cost, but continuing to sell a product that disappoints customers also has a cost. Junestrand frames the decision as a commitment to the product’s long-term quality.

The team developed a product manifesto to align its decisions around what users actually need. A shared standard can help teams make choices when many features are possible.

05

People and culture

He says building a company ultimately depends on people. His early hiring mistakes taught him that a prestigious résumé alone does not establish whether someone fits the work and the team.

He says building a company ultimately depends on people. His early hiring mistakes taught him that a prestigious résumé alone does not establish whether someone fits the work and the team.

He discusses culture and the signals leaders send through their own decisions. Values matter most when they influence who is hired, how disagreements are handled and what behavior is rewarded.

Junestrand connects ambition to his experience in Sweden and the social pressure sometimes described by the Law of Jante. He argues that a team can set high goals while remaining aware of its weaknesses.

He says the company tries to sustain urgency even after strong growth. His competitive framing is a management philosophy, not evidence that one culture works for every organization.

He discusses onboarding employees in Stockholm and exposing them to customer problems. Shared context can help people understand the product and work across functions.

Looking back, he emphasizes the accumulation of small actions over a single fundraising announcement or large customer win. Much of building happens through repeated problem solving.

06

Domain expertise and models

Asked whether domain expertise is essential, he says the founders learned from customers rather than beginning as lawyers. Their experience shows one path; legal expertise and professional review remain important for legal work.

Asked whether domain expertise is essential, he says the founders learned from customers rather than beginning as lawyers. Their experience shows one path; legal expertise and professional review remain important for legal work.

The speakers discuss building products while underlying models improve. Junestrand says the team tried to solve current customer problems while anticipating a near-term increase in model capability.

Betting on future model improvements involves risk: a promised improvement may arrive later than expected or fail on a particular task. A product still needs to deliver value with the systems available.

07

Evaluations and learning

They discuss building evaluations for legal workflows. A useful evaluation should reflect the real task and consequences of errors, rather than rely only on a general model leaderboard.

They discuss building evaluations for legal workflows. A useful evaluation should reflect the real task and consequences of errors, rather than rely only on a general model leaderboard.

Asked when someone should start a company, Junestrand emphasizes learning by doing and taking responsibility. He also recognizes that different people bring different circumstances and experience.

He reflects on the learning that comes from figuring out unfamiliar problems. Working in a startup before founding one can also provide experience; there is no single path through the decision.

08

Competition and ambition

The interview returns to competitiveness and goals. He wants teams to celebrate wins and learn from losses together, rather than treating success as an individual achievement.

The interview returns to competitiveness and goals. He wants teams to celebrate wins and learn from losses together, rather than treating success as an individual achievement.

On ambition, he describes how peers and environment can affect what founders believe is possible. Being around ambitious people can raise expectations, though it does not guarantee a particular outcome.

09

Communication and growth

He returns to communication as a founder skill: explaining the company’s mission to customers, employees and investors requires adapting the story to each audience without losing its substance.

He returns to communication as a founder skill: explaining the company’s mission to customers, employees and investors requires adapting the story to each audience without losing its substance.

The pair discuss technical skills, teamwork and the ability to learn. Junestrand argues that rapidly changing tools make a willingness to keep learning especially valuable.

He describes using agents inside company processes, including routing contract-related work. Such workflows require access controls and human checks appropriate to the task and its risks.

Asked about growth from $1 million to $100 million in annual recurring revenue, he describes rapid scaling and the value of learning from people who have done it before. The figure is a company-reported milestone, not audited evidence in the talk.

The closing exchange covers the company’s marketing and a campaign involving actor Jude Law. The episode ends with questions from the audience about building and communicating a growing company.

同じチャンネルの記事

スライド 7 枚1:05:44EN
Y Combinator· 2026年10月26日Marvin Schwaibold on AI-Powered Design, Interactive Prototypes and Better WebsitesAaron Epstein, Marvin Schwaibold · Y CombinatorShopify’s product design studio lead shows how he uses AI to build and refine interactions, then reviews six websites with Aaron Epstein.Product DesignAI interface designプロダクトと UX+1
スライド 32 枚1:50:48EN
Y Combinator· 2026年10月24日Why Inference Matters 12× More Than TrainingFrançois Chaubard, Dan Fu, Jeff Morgan, Joe Fioti, Liangsheng Yin · Y CombinatorYC Paper Club walks the inference stack from an API request down to GPU kernels. Four talks from Together AI, Ollama, Luminal, and SGLang argue that serving quality, not just training FLOPs, is where most of the compute and the money now sit.InferenceGPU programming & kernelsKV caching & prefix reuse+2
47:08EN
Y Combinator· 2026年10月23日How Farza Majeed Makes Product Demo VideosAaron Epstein, Farza Majeed · Y CombinatorA founder’s approach to frequent launches, storytelling and measuring the audience that converts.Company BuildingProduct Strategy
47:43EN
Y Combinator· 2026年10月22日Designing Beyond AI Defaults: Magic Patterns and Website ReviewsAaron Epstein, Alex Danilowicz · Y CombinatorAlex Danilowicz and Aaron Epstein discuss prototypes, context and clearer startup sites.Product DesignAI interface designStartups
1:37:07EN
Y Combinator· 2026年10月17日AI Safety as Engineering: Proofs, Stress Tests and Incident InvestigationClark Barrett, Mykel KochenderferStanford researchers discuss formal methods and safety-critical evaluation at YC Paper Club.AI SafetyCode review & verification評価(Evals)
43:23EN
Y Combinator· 2026年10月8日Jeetu Patel on Cisco’s AI Infrastructure Bet and Engineering CultureJared Friedman, Jeetu PatelThe Cisco executive talks with YC’s Jared Friedman about growth, security and career lessons.エンタープライズ AIインフラとデプロイLeadership