AI Legal Software with Scott Stevenson
Episode
56 min
Read time
2 min
Topics
Investing, Startups, Design & UX
AI-Generated Summary
Key Takeaways
- ✓Product-Market Fit Testing: Run experiments in three sequential stages before scaling: first test whether a landing page captures email signups, then whether users pay, then whether they return repeatedly. Most ideas fail at stage one. Retention is the only metric that cannot be faked by strong salespeople or generous contract terms.
- ✓AI-First vs. AI-Added: Launching a standalone AI product outperforms adding AI as a feature to an existing platform. Spellbook was rebuilt from scratch rather than appended to Rally, which prevented the new capability from being buried as item 20 of 20 in a feature list and allowed a fundamentally different user experience.
- ✓Hallucination Mitigation via Optionality: Present AI output as a menu of choices rather than a single authoritative answer. Spellbook surfaces roughly 100 suggested contract changes per review; lawyers accept approximately 50% of them. This removes the need for perfect accuracy and reframes the tool as a thread-finder rather than a decision-maker.
- ✓Custom Model Timing: Avoid investing heavily in fine-tuned or proprietary models prematurely. Spellbook's initial version was largely a GPT-3 wrapper with prompt engineering. Because foundation models improve rapidly, resources spent training narrow models can be rendered obsolete by the next public release, wasting capital that early-stage startups cannot afford.
- ✓Workflow-Embedded UX: Build AI tools directly inside the software users already operate rather than creating a separate chat interface. Spellbook runs inside Microsoft Word because lawyers refuse to leave it. Embedding suggestions within existing workflows increases adoption and preserves the user's sense of control, which is especially critical in high-stakes professional environments.
What It Covers
Scott Stevenson, co-founder of Spellbook, describes building an AI contract review and drafting tool for commercial lawyers. Starting as Rally in 2019, the company ran nearly 100 product experiments before launching its LLM-based copilot in 2022, now serving close to 2,000 paying law firms.
Key Questions Answered
- •Product-Market Fit Testing: Run experiments in three sequential stages before scaling: first test whether a landing page captures email signups, then whether users pay, then whether they return repeatedly. Most ideas fail at stage one. Retention is the only metric that cannot be faked by strong salespeople or generous contract terms.
- •AI-First vs. AI-Added: Launching a standalone AI product outperforms adding AI as a feature to an existing platform. Spellbook was rebuilt from scratch rather than appended to Rally, which prevented the new capability from being buried as item 20 of 20 in a feature list and allowed a fundamentally different user experience.
- •Hallucination Mitigation via Optionality: Present AI output as a menu of choices rather than a single authoritative answer. Spellbook surfaces roughly 100 suggested contract changes per review; lawyers accept approximately 50% of them. This removes the need for perfect accuracy and reframes the tool as a thread-finder rather than a decision-maker.
- •Custom Model Timing: Avoid investing heavily in fine-tuned or proprietary models prematurely. Spellbook's initial version was largely a GPT-3 wrapper with prompt engineering. Because foundation models improve rapidly, resources spent training narrow models can be rendered obsolete by the next public release, wasting capital that early-stage startups cannot afford.
- •Workflow-Embedded UX: Build AI tools directly inside the software users already operate rather than creating a separate chat interface. Spellbook runs inside Microsoft Word because lawyers refuse to leave it. Embedding suggestions within existing workflows increases adoption and preserves the user's sense of control, which is especially critical in high-stakes professional environments.
Notable Moment
Stevenson admitted he launched Spellbook expecting to eliminate lawyers entirely, only to conclude years later that contract law is objectively harder than software engineering — largely because signed legal documents are nearly irreversible, whereas code errors can be rolled back almost instantly.
Episode Transcript
Welcome to the Axial podcast. Axial is an early stage investment firm based in San Francisco. We partner with great founders and inventors investing in early stage life science companies often when they are no more than an idea. Axial is fanatical about helping the right venture who's compelled to build their own and journey business. Okay, Scott. Thanks for joining this podcast. I'm really excited to talk about SvelteBook and also your story of how you got here. But maybe to start off, you can just what does SvelteBook do and, what's the mission? Yep. Thanks thanks for having me, Josh. So Spellbook is, really an AI copilot, for primarily commercial lawyers who are working on, contracts and, contract like documents. And, it enables them to draft and review these types of documents up to 10 times faster, using a variety of kind of, LLM, large language model based tools or generative AI kind of tools. So for example, something you can do is if, a lawyer has a 100 page contract that they're asked to review by their client, we can actually review based on their instructions and, you know, find, a 100 things that need to be changed in that contract to help protect the client or make the deal more preferable for them or, you know, fix issues and oversights and things like that. And so that's, essentially what we do. We service other types of lawyers as well, other family lawyers and and litigators, as well. But, you know, primarily, we work on these sorts of, corporate and commercial transactional docs. And, yeah, that's us. And so so for spell book on this contract review drafting, why has it been a great market for you to grow in? Do you have, like, maybe, I think, 2,000, lawyers using the product? And how has the progress in AI over the last, let's say, two years enabled better features or even new features for your customers? Yeah. Yeah. So we've had, yeah, almost a 100,000 people sign up one way or another to to try out our product. And Wow. You know, to today today, we have almost 2,000 law firms, you know, paying for it active actively using it, in terms of, like, you know, qualified law firms, in many cases with multiple lawyers, at those firms. So there's been a lot of interest in we've kind of been letting letting the interest in, based on, you know, we've only been selling it to, qualified lawyers, licensed lawyers, and things like that. I know. I honestly, I was trying to I was trying to use guys a product, and I couldn't even sign up. I was like, how do I use this product? And it's just that you guys I bet you have more way more interest than what's on the surface. Yeah. It's it's it's it's your product is actually really hard to use unless you're, like, a qualified lawyer. So Yeah. Yeah. That's something …
Get the full transcript (10,183 words) + summary by email — free
One-time email with the complete transcript and AI summary of this episode. No account needed.
One email, no spam. We’ll also show you what SignalCast does.
You just read a 3-minute summary of a 53-minute episode.
Get Axial Podcast summarized like this every Monday — plus up to 2 more podcasts, free.
Pick Your Podcasts — FreeKeep Reading
More from Axial Podcast
Modern Computational Tools for Chemistry with Corin Wagen
Mar 23 · 50 min
All-In with Chamath, Jason, Sacks & Friedberg
Saronic Founders: Autonomous Warships, China's 230X Advantage & Swarms of Robot Ships
Aug 5
More from Axial Podcast
Evolutionary Intelligence and Biologics Discovery with Jeremy Agresti
Mar 23 · 51 min
The Prof G Pod
Why OpenAI Bought a Podcast — with TBPN’s John Coogan and Jordi Hays
Jul 9
Books, tools, and gear mentioned in this episode
SignalCast may earn commission on purchases via these links. As an Amazon Associate, SignalCast earns from qualifying purchases.
More from Axial Podcast
We summarize every new episode. Want them in your inbox?
Modern Computational Tools for Chemistry with Corin Wagen
Evolutionary Intelligence and Biologics Discovery with Jeremy Agresti
AI Workflows for Biopharma with Alex Telford
Scaling Proteomics with Milad Dagher
Proteomics and AI with Peter Cimermančič
Similar Episodes
Related episodes from other podcasts
All-In with Chamath, Jason, Sacks & Friedberg
Aug 5
Saronic Founders: Autonomous Warships, China's 230X Advantage & Swarms of Robot Ships
The Prof G Pod
Jul 9
Why OpenAI Bought a Podcast — with TBPN’s John Coogan and Jordi Hays
The Knowledge Project
May 12
Winston Weinberg: Speed, Stress, and Better Decisions
Invest Like the Best with Patrick O'Shaughnessy
Apr 14
Scott Nolan - SpaceX, Founders Fund, and Rebuilding American Uranium Enrichment - [Invest Like the Best, EP.467]
How I Built This
Feb 23
Square: Jim McKelvey. He Lost a $2,000 Sale, Then Built a $10 Billion Company
Explore Related Topics
This podcast is featured in Best Biotech Podcasts (2026) — ranked and reviewed with AI summaries.
Read this week's Investing & Markets Podcast Insights — cross-podcast analysis updated weekly.
You're clearly into Axial Podcast.
Every Monday, we deliver AI summaries of the latest episodes from Axial Podcast and 192+ other podcasts. Free for one show.
Start My Monday DigestNo credit card · Unsubscribe anytime