I open-sourced my legal RAG platform, would love you to use it with local models
I’ve been building a legal RAG platform called CaseCite and decided to open-source the entire project.
The idea is pretty simple, give legal pros the ability to run AI against their own documents, on their own infrastructure, without expensive software subscriptions or giving up control of their data.
It lets you upload and search your document library, chat with your documents, and get answers with citations linking directly to the source material.
I’ve also built in contract analysis and redlining, legal drafting, case-law research through CourtListener, citation checking, and judge intelligence.
It’s fully self-hosted, supports bring-your-own API keys, and can run with locally hosted models. The code is released under Apache 2.0, so you’re free to use and modify it.
It’s not trying to replace every feature of an enterprise legal platform, but I think it covers a lot of what smaller firms or solo attorneys need.
Would love feedback from attorneys, developers, or anyone working in legal tech.
thomasunise/casecite
41 points · 10 comments
📰 Hacker News (AI)
Harvey raised 500M for legal AI, I built the same and open sourced it
thomasunise/casecite
#IA #AI #ML
I open-sourced my legal RAG platform for law firms, would love you to use it
I’ve been building a legal RAG platform called CaseCite and decided to open-source the entire project.
The idea is pretty simple, give law firms the ability to run AI against their own documents, on their own infrastructure, without expensive software subscriptions or giving up control of their data.
It lets you upload and search your document library, chat with your documents, and get answers with citations linking directly to the source material.
I’ve also built in contract analysis and redlining, legal drafting, case-law research through CourtListener, citation checking, and judge intelligence.
It’s fully self-hosted, supports bring-your-own API keys, and can run with locally hosted models. The code is released under Apache 2.0, so you’re free to use and modify it.
It’s not trying to replace every feature of an enterprise legal platform, but I think it covers a lot of what smaller firms actually need.
Would love feedback from attorneys, developers, or anyone working in legal tech.
thomasunise/casecite
34 points · 21 comments
Harvey raised 500M for legal AI, I built the same and open sourced it
Following Harvey's $500M raise for legal AI, an open-source alternative has been released on GitHub: thomasunise/casecite
5 views
Harvey raised $500M for legal AI. Someone just built the same thing and open-sourced it: CaseCite, self-hosted legal RAG for caselaw research and document analysis. The moat is distribution, not the model. thomasunise/casecite
4 views
Harvey raised 500M for legal AI, I built the same and open sourced it
thomasunise/casecite
Harvey raised 500M for legal AI, I built the same and open sourced it
69 points · 38 comments
@inference_net Do you have an api?
Would love to add it to my project
thomasunise/argus
1 like · 416 views
Argus keeps agents read-only, routing writes through a broker for policy, approvals, dry runs and rollback. Collectors lack live-account tests and 12 rules are untuned, so monitor-only rollout makes sense.
thomasunise/argus
1 like · 128 views
Mentions
I open-sourced my legal RAG platform, would love you to use it with local models I’ve been building a legal RAG platform called CaseCite and decided to open-source the entire project. The idea is pretty simple, give legal pros the ability to run AI against their own documents, on their own infrastructure, without expensive software subscriptions or giving up control of their data. It lets you upload and search your document library, chat with your documents, and get answers with citations linking directly to the source material. I’ve also built in contract analysis and redlining, legal drafting, case-law research through CourtListener, citation checking, and judge intelligence. It’s fully self-hosted, supports bring-your-own API keys, and can run with locally hosted models. The code is released under Apache 2.0, so you’re free to use and modify it. It’s not trying to replace every feature of an enterprise legal platform, but I think it covers a lot of what smaller firms or solo attorneys need. Would love feedback from attorneys, developers, or anyone working in legal tech. thomasunise/casecite
📰 Hacker News (AI) Harvey raised 500M for legal AI, I built the same and open sourced it thomasunise/casecite #IA #AI #ML
I open-sourced my legal RAG platform for law firms, would love you to use it I’ve been building a legal RAG platform called CaseCite and decided to open-source the entire project. The idea is pretty simple, give law firms the ability to run AI against their own documents, on their own infrastructure, without expensive software subscriptions or giving up control of their data. It lets you upload and search your document library, chat with your documents, and get answers with citations linking directly to the source material. I’ve also built in contract analysis and redlining, legal drafting, case-law research through CourtListener, citation checking, and judge intelligence. It’s fully self-hosted, supports bring-your-own API keys, and can run with locally hosted models. The code is released under Apache 2.0, so you’re free to use and modify it. It’s not trying to replace every feature of an enterprise legal platform, but I think it covers a lot of what smaller firms actually need. Would love feedback from attorneys, developers, or anyone working in legal tech. thomasunise/casecite
Harvey raised 500M for legal AI, I built the same and open sourced it
Following Harvey's $500M raise for legal AI, an open-source alternative has been released on GitHub: thomasunise/casecite
Harvey raised $500M for legal AI. Someone just built the same thing and open-sourced it: CaseCite, self-hosted legal RAG for caselaw research and document analysis. The moat is distribution, not the model. thomasunise/casecite
Harvey raised 500M for legal AI, I built the same and open sourced it thomasunise/casecite
Harvey raised 500M for legal AI, I built the same and open sourced it
@inference_net Do you have an api? Would love to add it to my project thomasunise/argus
Argus keeps agents read-only, routing writes through a broker for policy, approvals, dry runs and rollback. Collectors lack live-account tests and 12 rules are untuned, so monitor-only rollout makes sense. thomasunise/argus