The boring genius of how Atrium kills legal busy work | Hacker News Reader

The boring genius of how Atrium kills legal busy work

techcrunch.com

108 pointswalterbell33 comments

Our startup signed on with Atrium and they've been superb to work with. The attorneys are great, the pricing is reasonable, and they're very fast at turning around documents. I never appreciated just how much speedy turnaround can make the legal process more pleasant.

I don't think they've quite hit their stride yet with the add-on services. I agree with the comments so far: It's not that difficult to organize your trove of legal documents in Google Docs, and I'd rather control my own hiring doc templates with a general-purpose tool like DocuSign than add another tool in the mix.

But the bigger picture here is the fact that they're releasing any such services at all. The chance of a traditional law firm releasing these services is zero, and there are software-based services I really want.

Personally, I'm waiting for the tool where I can declare my company policies around NDAs, and then use Atrium NDA (or whatever) to automatically scan the NDA I received, tell me how it violates my policies, and auto-generates a response to the counterparty.

Accommodating pointless diversity of NDAs is enabling dysfunction. Have a look at https://waypointnda.com/.

'DocuSign template .zip' link is broken.

Thanks for mentioning! I’ll fix it.

Has this been reviewed by lawyers outside the U.S.? Section 6, the 18 U.S.C. 1833(b) notice seems extremely specific, and I'm worried that there might be other subtleties that make this unusable outside the U.S. and/or other required items for the NDA to function outside the U.S.

(The U.S.-centrism in the software industry has always been a pet peeve of mine.)

Yes. I’m a Canadian lawyer who has corresponded with the author of that NDA. I pointed out the same language and he’s working on a version that’ll account for other countries. The author is keenly aware of this and wants to make it not US-centric.

Howdy, a3camero!

Indeed, we're currently working on a version 2.0.0. The new release will implement a few small, general improvements, and well as feedback from counsel from a few common-law jurisdictions.

beefhash, if you know lawyers from other jurisdictions, please put them in touch with me at kyle@kemitchell.com. I'd love to provide them the latest working draft of 2.0.0 and read their thoughts.

Does your experience with Atrium suggest that this kind of functionality is feasible? Even developing a 'law as code' platform where NDAs can be written by computer experts in a DSL, formally analysed and compiled to English seems like a major engineering challenge, let alone building a system that can accept an unstructured PDF and parse out its legal implications. That's a far cry from the 'busy work' (filling out templates, making sure the right people have signed them, and keeping copies) discussed in this article.

At this point in our relationship, Atrium's basically been a really well-run law firm who is economically incentivized to make legal more efficient. They also have significant funding, an experienced CEO, and silicon valley thinking, so yeah I think the foundation for something significant like "law as code" is there, even if it's years away.

That being said, I agree the first releases are underwhelming but "If you're not embarrassed by the first version of your product, you've launched too late." [1]

[1] https://twitter.com/reidhoffman/status/847142924240379904?la...

Legal English documents are already a "DSL" to non-experts.

With recent advances on Neural Network models such as GPT-2, we (engineers, researchers, etc) are closer to be able to analyse any documents written in correct English. It won't be fully autonomous at first, but we can imagine it will help to answer questions such as "given my policies database, do I break the rules defined in that contract?". The tool will give hints, which will be reviewed by an actual lawyer to say the final word.

It is the same idea than https://books.google.com/talktobooks/ , but for legal

I assume they are just figuring out which PDFs from different companies contain the same language and then letting their lawyers compare them

Technology enabled service companies have the potential to be one of the next big waves in software. I spent last summer trying to figure out what I wanted to do for my next startup and it was amazing to me how many businesses run on paper, excel spreadsheets and email. Trucking, recruiting, cleaning, building security, building maintenance, whatever.

The beauty of a tech enabled service company is that you can build it without raising outside funding. You use the earnings from the services to fund product development.

The other great thing is that the market is already proven. You just build a competitive advantage through software that allows you to work more efficiently or provide a better customer experience.

I ended up starting a technology enabled staffing agency for the tech industry.

The industries you mention are huge. And fragmented.

But once we start adding technology(not necessarily just software), they will become much less fragmented, with a few big winners.

Doesn't that mean venture capital is a necessity ?

No.

Maybe. I don't think Atrium will ever become one of 2 or 3 big law firms. There will always be thousands of business law firms. Tech enabled service companies don't seem to have the same winner-take-all nature you might see in platform or network effects businesses.

Even without the winner take all potential, it still makes for a great $100M business idea that seems much less risky to me. There can be dozens or hundreds of $100M trucking companies. That's a lot of winners. :)

My suspicion is that AI/ML will drastically change the mechanics of how this works. Down the line, if Atrium has access to a million documents that it can compare and contrast to, it might be significantly better then the legal service with none.

The worst bit is that this will cause a reinforcing loop, of more customers flooding to the market leader, causing the market leader to develop better AI, causing more customers, etc etc.

Let's say I could offer jobs to every truck accross the US, and had access to every trucking job customer.

This enables me to plan the best routes and be the most efficient. My platform will win.

But you imply there are hundreds of niches.

That's interesting.

What am I missing ? How do you analyze that market ?

Relationships matter. Independent truckers or small trucking companies have close relationships with their customers. The matters in both the Business-As-Usual case (Ron drives a beer distribution truck which makes the same route to the same bars twice a week) and the Not Normal case (Hey Fred, I have a load that needs special handling and immediate delivery response).

Large, generic haulers just don't offer the same kind of customer service.

Atrium presents itself as a law firm, even though it is two different entities. One of the entities is a law firm, the other is a tech startup. From their web site, in light gray on white background:

> "Atrium LLP is a registered limited liability partnership under the laws of the State of California. Atrium Legal Technology Services, Inc. is a non-law corporation incorporated under the laws of the State of Delaware."

Aren't there some legal ethics problems there? Is Atrium the law firm sharing fees with non-lawyers? Is Atrium LTS engaging in the unlicensed practice of law?

Can't talk about ethics, but I'd imagine a group of lawyers have put some thought into the legal aspects of their company.

Fees and info get shared with non-lawyers all the time. As long as the right contracts are in place, there's nothing problematic about it. If Atrium LTS structures the decisionmaking so that Atrium LLP is giving the legal advice for the purposes of the law, then there's nothing shaky here. At some point this will all get hashed out in court a few dozen times, but there's nothing about this that looks insoluble.

Yes, lawyers pay non-lawyers for services all time. I’m thinking of profit-sharing. We don’t know how close to profit-sharing their arrangement is, because we don’t know the details of the contract between them.

I’m sure their lawyers have good arguments that their business is free of conflicts of interest. However, that they appear to intentionally blur the line between the company and the firm is suspicious.

What do you find suspicious about it? You've acknowledged that they're presumably taking care to comply with the letter of the law. As long as legal advice is being provided by lawyers whose ethical duties override their personal interests (eg. share holdings in a closely-related non-law company), what's the problem? Other jurisdictions have relaxed the rule against profit-sharing and I don't think there have been any major ethical problems as a result.

The big 4 accounting firms do a huge amount of legal work for clients. It’s done by people who have passed the bar but are “not practicing attorneys” at the firm.

There's an entire industry devoted to the idea of that profit-sharing (legal/litigation financing), and many large firms are financed by what is essentially a litigation investment company. Very little of this is new ground.

As far as I know, litigation finance is just a high-interest, non-recourse loan secured solely by a future recovery from a lawsuit. It is used by plaintiffs to pay for personal expenses during a lawsuit. Plaintiffs just promise to repay the principal amount plus interest in case of a recovery, it's not an assignment of all or a portion of their interest in the recovery.

So, maybe I misunderstand this. Are law firms actually selling a percentage of their fee in contingent cases to these companies? Or is it structured as a high-interest loan like the loans to plaintiffs?

One of the scenarios they describe - hiring documents - strikes me as a VERY limited value proposition. I can’t imagine someone paying a lawyer to review their hiring documents more than once. All of the associated documents (offer letter, non-disclosure, non-compete, assignment, stock options, etc.) are so easily automated with a basic eSignature tool like HelloSign. Our attorneys simply gave us template docs (billing very little time) and we’ve used them for dozens of employees and multiple due diligence processes. The actual ROI sounds minimal at best.

I can see this being more useful in a sales situation where you need to make concessions much more regularly and need flexibility in your legal documents. Depending on how hard Atrium is to configure, I suspect it could be a lower cost version of setting up Salesforce CPQ. However, you likely still want this process to be integrated into Salesforce at some point so you’ll have to do that separately still.

Also, the frequency of automating your financing docs is not very compelling. In my experience, your investors provide the term sheets and then you need to redline them back to them. The idea of scanning these types of documents to extract the meaningful data points is very cool (ie. Comparing valuations automatically across spreadsheets) but I would still say a manual review from an attorney is essential due to the risk/reward of missing some small item that isn’t conventional in a term sheet.

I’m curious if anyone here has used Atrium and has see a compelling ROI for your team?

Yeah. Docusign is a little annoying but we also have the 5 variables (name/title/salary/options/start date) pretty well templatized.

When I heard about Atrium's fundraising, I assumed they had shown some pretty sophisticated ML under the hood (or at least a plan to do so).

I thought that "conventional" platform tech - basically, document management - wouldn't be compelling enough for venture capital investment.

I guess I was wrong? There's nothing here that looks particularly unique. Not saying there's no demand, but I can't understand where the 1000x returns are going to come from.

Are people really paying lawyers to fill out offer letters and the like? My lawyers in the early 1990s would just send me templates to fill out use -- for free.

There must be more to this than is in the article.

This reads like an ad, and doesn’t belong on the HN front page.

I’ve been using Clerky for awhile, which is a little clunky but certainly can be useful.

I’m curious how these two platforms compare. The common use cases seem to have a lot of overlap.

If I were a client of a lawyer, I would never allow them to upload my data to a third party service.

Reply on news.ycombinator.com