jacobheller
Joined 13 years ago
Karma
232
Submissions
83
Member since
Mar 4
CEO, Casetext (YC S13) https://www.casetext.com
View on Hacker NewsRecent submissions
Comment · 2 years ago
Very slick and potentially very powerful. After a few minutes playing with it, I have a few recommendations:
- Variables should have more types, like an array of objects
- Prompting should incorporate Jinja2/Nunjucks
- For every prompt, I should be able to create many different test examples, along with an answer key, and measure how well it does across many tests
- It should auto-save. I did a lot of prompting work and then clicked another icon. When I came back, all my work was gone. (In fact, I don't see where to save at all! Maybe I'm just missing it.)
Comment · 9 years ago
Casetext | Software Engineer, Research Scientist, Machine Learning Engineers, and More | San Francisco, CA | ONSITE
This is an insanely fun time to join us. If you are interested in search, NLP, recommender engines, cutting-edge front-end technologies, or making a difference through your work, you should check us out.
Our open positions are here: https://jobs.lever.co/casetext/
Comment · 9 years ago
Our startup, Casetext (YCS13) was mentioned a few times here, so I thought I'd stop in.
The crux of the article is that most legal research solutions have ignored the immense power contained in the links between laws:
> Laws frequently reference other laws in order to reuse definitions, introduce exceptions, or make it clear that two concepts are meant to work together. Consequential laws tend to get referenced in other laws as their influence spreads throughout the legal system. Experienced lawyers build up detailed mental maps of these links, allowing them to jump immediately to core issues of complex legal problems.
> However, most laws can only be searched using the dark-age, Lycos strategy—guess at keywords and hope—and it’s often necessary to pay for even that limited functionality.
We at Casetext are taking a very different approach than the "dark-age, Lycos strategy" that you have to pay for:
1. On Casetext, the law is free, as is basic search. Honestly, it's insane that Westlaw and LexisNexis charge as much as they do for basic keyword search over a database that should have been free to begin with.
2. We make money by charging for advanced, data-driven ways that lawyers can research more efficiently. CARA, our premium product, enables a lawyer to drag-and-drop upload a document they're working on, and will recommend the research that the lawyer missed but is very relevant to what they're working on (https://casetext.com/cara). A key ingredient behind this awesome tech is the network of citations that the article mentions.
Whether it's us or other startups, I agree with the article that in the next few years you'll see a trend towards more "Google for Law" -- companies will make legal research free, and their comparative advantage will be on their technology, often driven by ML/AI. As a lawyer/coder, it's a pretty exciting time to be in the space.
Oh yeah, and we're hiring! https://casetext.com/jobs