3 ms·

I'm building https://uscodex.org https://uscodex.org. In a nutshell, the goal is to collect all US federal law in version control, and cross-link any references

I'm building https://uscodex.org https://uscodex.org. In a nutshell, the goal is to collect all US federal law in version control, and cross-link any references between different sections. Currently I have the US Code, CFR, ~30 years of public laws and congressional bills and 25 years of executive orders. Actually using raw git repos as the main data store which has worked out great since the US Code changes very little between new releases and so it compresses very well. Also makes it easy to generate diffs of sections: https://uscodex.org/usc/d/119-100/119-102not101/16/2201 https://uscodex.org/usc/d/119-100/119-102not101/16/2201

hdb2 19d ago
this seems like an outstanding idea!
- SebRollen 19d ago
  Aw, thanks! I thunk it myself :) The cross-referencing between different corpora has actually been very helpful. I was able to find a typo in the notes to a US Code section through it and submitted it to the OLRC. They fixed it, so now I will forever be bragging about having fixed an error in the US Code haha
  - pulpconversatio 16d ago
    Well done on this! Two questions for you. 1) How much do you know about Policy Genius? Perhaps their trajectory could be instructive in what went well / wrong? 2) What do you think about adding a search for public equities? Type in AAPL, NVDA, etc.. and then UScodex delivers a code analysis that could surface relevant regulatory milestones / capture. If it works this could be a professional insights tool and/or "risk and compliance" software. Good luck.
    - pulpconversatio 16d ago
      Oops - I meant FiscalNote not Policy Genius. Sorry.
epiccoleman 18d ago
Hoo boy. It's an awesome idea, don't get me wrong - and it's work that should definitely be done. But I've worked professionally in this space and it is wildly complicated. I've never seen how things are done on the federal side, but US states have various modes of "electronic filing" for rules and codes - and there's a ton of complexity. It's the kind of thing that sounds insanely boring on the surface, but there's a lot of interesting problems to solve, lots of coordination between various parties, lots of "deep state"[0] politicking - and so it becomes a surprisingly cool area to learn about and work in, and really matters in a way that many software projects do not. Best of luck - it's a cool problem space. [0]: I use "deep state" here somewhat ironically, but also just in the literal definitional sense: "people who are not elected but who work for the government and are surprisingly important". 0% intent to use it in the often insulting manner of popular discourse.
- SebRollen 18d ago
  Thanks! Super interested in hearing more about some of the battle-scars you've acquired from working in the space if you're open to sharing? Very early in the journey here, so still have a ton to learn! So far, it seems like the federal space is relatively well-behaved compared to your experience with states. Most agencies standardized on the USLM schema[^1][^2] in 2013. Before that, it's a hodge-podge of XML and plain-text sources. Also, FWIW, my interactions with federal agencies so far have been insanely positive. When I reported the error I found in the US code at 9pm, a staffer from the OLRC verified the issue and got back to me by 7am the next day, more efficient than most private sector organization's I've interacted with! [^1]: https://github.com/usgpo/uslm https://github.com/usgpo/uslm. [^2]: Meta: The USLM itself was created by federal law https://uscodex.org/usc/2/181 https://uscodex.org/usc/2/181
  - epiccoleman 18d ago
    I am not sure how much I'm allowed to say publicly between NDAs and contracts and all that fun stuff (I may already be over the line, tbh) - but feel free to drop me an email at the address in my profile and I can point you to a few things that might be of interest.
- 18d ago
andriy_koval 18d ago
project is cool, but what you will do with it? Do you have audience, marketing strategy, monetization plans, etc?..
- SebRollen 18d ago
  Thanks! The primary goal is to make it the best place to research US federal law. All the features I have currently (cross-corpus search, version diffing, backlinks, law lineage, etc) will stay free and open. Audience is primarily lawyers and legal researchers in fields that rely heavily on statutory law (tax, bankrupty, immigration, etc). I'm blessed to have a plethora of lawyers in my family/friends, so planning on using them as ~guinea pigs~ beta testers. If you (or anyone else reading this) think this is something you might use, I would love to talk more! seb@uscodex.org In terms of monetization, I'm not really sure yet. My hosting costs are very cheap at the moment, so I'm not in any rush there, but I'm imagining I'll build in some more premium features like an API (for other legal-techs), change monitoring (for anyone relying on specific statutes in their work) or team research features (shared bookmarks, notes, collections etc). Open to suggestions here too! One very cool thing about my data pipeline is I've essentially built a connected graph of federal law with tens of millions of edges across millions across hundreds of thousands of sections, and I think there will be a lot of powerful applications of that dataset
  - andriy_koval 18d ago
    > a connected graph of federal law with tens of millions of edges across millions across hundreds of thousands of sections, and I think there will be a lot of powerful applications of that dataset I am mostly curious because I built similar graph, but in different domain, so I am wondering what are the applications and what would be go to market strategy.
    - SebRollen 18d ago
      Yeah, it's a good question, let me know if you figure out a good answer :) Honestly, my focus right now is to make the site useful rather than profitable. This is a side-project that spawned out of needing something like this for my own research, so haven't figured out how/if I'll monetize it yet. The one idea I have related to the graph structure that I think could be valuable right now is letting users "subscribe" to particular sections of the law, with alerts for when there are new bills in Congress that might affect or alter it. The graph currently goes from bills > public laws > USC > CFR, so users could monitor a regulation and get notified as a bill moves its way through congress
zackmorris 17d ago
This is great! We need something like git blame to show how legislation evolved over time, who passed it, which lobbyists pushed for or wrote it, who benefited by receiving more money/power, and who had to foot the bill. This is becoming urgent, now that there's almost no correlation between what laws get passed and their popularity with the public, since Congress mainly only listens to moneyed interests: https://www.youtube.com/watch?v=5tu32CCA_Ig https://www.youtube.com/watch?v=5tu32CCA_Ig Discussion: https://act.represent.us/sign/problempoll-fba https://act.represent.us/sign/problempoll-fba By making corruption more transparent, we can make it so toxic that people participating in it get investigated and brought up on charges. This will become inevitable due to the arrival of AI, but speeding up its delivery is paramount if we want to save our country.

hdb2 19d ago

this seems like an outstanding idea!

SebRollen 19d ago
Aw, thanks! I thunk it myself :) The cross-referencing between different corpora has actually been very helpful. I was able to find a typo in the notes to a US Code section through it and submitted it to the OLRC. They fixed it, so now I will forever be bragging about having fixed an error in the US Code haha
- pulpconversatio 16d ago
  Well done on this! Two questions for you. 1) How much do you know about Policy Genius? Perhaps their trajectory could be instructive in what went well / wrong? 2) What do you think about adding a search for public equities? Type in AAPL, NVDA, etc.. and then UScodex delivers a code analysis that could surface relevant regulatory milestones / capture. If it works this could be a professional insights tool and/or "risk and compliance" software. Good luck.
  - pulpconversatio 16d ago
    Oops - I meant FiscalNote not Policy Genius. Sorry.

epiccoleman 18d ago

Hoo boy. It's an awesome idea, don't get me wrong - and it's work that should definitely be done. But I've worked professionally in this space and it is wildly complicated. I've never seen how things are done on the federal side, but US states have various modes of "electronic filing" for rules and codes - and there's a ton of complexity. It's the kind of thing that sounds insanely boring on the surface, but there's a lot of interesting problems to solve, lots of coordination between various parties, lots of "deep state"[0] politicking - and so it becomes a surprisingly cool area to learn about and work in, and really matters in a way that many software projects do not. Best of luck - it's a cool problem space. [0]: I use "deep state" here somewhat ironically, but also just in the literal definitional sense: "people who are not elected but who work for the government and are surprisingly important". 0% intent to use it in the often insulting manner of popular discourse.

SebRollen 18d ago
Thanks! Super interested in hearing more about some of the battle-scars you've acquired from working in the space if you're open to sharing? Very early in the journey here, so still have a ton to learn! So far, it seems like the federal space is relatively well-behaved compared to your experience with states. Most agencies standardized on the USLM schema[^1][^2] in 2013. Before that, it's a hodge-podge of XML and plain-text sources. Also, FWIW, my interactions with federal agencies so far have been insanely positive. When I reported the error I found in the US code at 9pm, a staffer from the OLRC verified the issue and got back to me by 7am the next day, more efficient than most private sector organization's I've interacted with! [^1]: https://github.com/usgpo/uslm https://github.com/usgpo/uslm. [^2]: Meta: The USLM itself was created by federal law https://uscodex.org/usc/2/181 https://uscodex.org/usc/2/181
- epiccoleman 18d ago
  I am not sure how much I'm allowed to say publicly between NDAs and contracts and all that fun stuff (I may already be over the line, tbh) - but feel free to drop me an email at the address in my profile and I can point you to a few things that might be of interest.
18d ago

andriy_koval 18d ago

project is cool, but what you will do with it? Do you have audience, marketing strategy, monetization plans, etc?..

SebRollen 18d ago
Thanks! The primary goal is to make it the best place to research US federal law. All the features I have currently (cross-corpus search, version diffing, backlinks, law lineage, etc) will stay free and open. Audience is primarily lawyers and legal researchers in fields that rely heavily on statutory law (tax, bankrupty, immigration, etc). I'm blessed to have a plethora of lawyers in my family/friends, so planning on using them as ~guinea pigs~ beta testers. If you (or anyone else reading this) think this is something you might use, I would love to talk more! seb@uscodex.org In terms of monetization, I'm not really sure yet. My hosting costs are very cheap at the moment, so I'm not in any rush there, but I'm imagining I'll build in some more premium features like an API (for other legal-techs), change monitoring (for anyone relying on specific statutes in their work) or team research features (shared bookmarks, notes, collections etc). Open to suggestions here too! One very cool thing about my data pipeline is I've essentially built a connected graph of federal law with tens of millions of edges across millions across hundreds of thousands of sections, and I think there will be a lot of powerful applications of that dataset
- andriy_koval 18d ago
  > a connected graph of federal law with tens of millions of edges across millions across hundreds of thousands of sections, and I think there will be a lot of powerful applications of that dataset I am mostly curious because I built similar graph, but in different domain, so I am wondering what are the applications and what would be go to market strategy.
  - SebRollen 18d ago
    Yeah, it's a good question, let me know if you figure out a good answer :) Honestly, my focus right now is to make the site useful rather than profitable. This is a side-project that spawned out of needing something like this for my own research, so haven't figured out how/if I'll monetize it yet. The one idea I have related to the graph structure that I think could be valuable right now is letting users "subscribe" to particular sections of the law, with alerts for when there are new bills in Congress that might affect or alter it. The graph currently goes from bills > public laws > USC > CFR, so users could monitor a regulation and get notified as a bill moves its way through congress

zackmorris 17d ago

This is great! We need something like git blame to show how legislation evolved over time, who passed it, which lobbyists pushed for or wrote it, who benefited by receiving more money/power, and who had to foot the bill. This is becoming urgent, now that there's almost no correlation between what laws get passed and their popularity with the public, since Congress mainly only listens to moneyed interests: https://www.youtube.com/watch?v=5tu32CCA_Ig https://www.youtube.com/watch?v=5tu32CCA_Ig Discussion: https://act.represent.us/sign/problempoll-fba https://act.represent.us/sign/problempoll-fba By making corruption more transparent, we can make it so toxic that people participating in it get investigated and brought up on charges. This will become inevitable due to the arrival of AI, but speeding up its delivery is paramount if we want to save our country.