Original Summary

I’ve been building a legal-tech project for a while and would really like some outside feedback on it I've been working on something called LAWNOVA for a while now. It started fairly small, but over time it has turned into a much bigger experiment around a question I've been thinking about: What would legal software look like if AI wasn't just another chatbot sitting on top of it? There are obviously a lot of AI legal products appearing now, and things seem to be moving especially fast lately. CourtListener is making case-law infrastructure much easier to work with programmatically, specialized legal AI is becoming a real category, and the tooling available to small teams is completely different from what it was even a year or two ago. So we've been building and testing our own approach. It's already running as a pilot, but it's still very much something we're improving, breaking, rebuilding and learning from. The main platform is: LAWNOVA https://lawnova.pro The idea is eventually to connect the pieces of actual legal work instead of having a collection of unrelated AI tools: cases, documents, research, transcripts, evidence, drafting, trial preparation, billing, communication and so on. We also have a couple of more focused tools that came out of that work. PDF / Transcript Research https://pdf.lawnova.pro This one is built around searching and working with large legal PDFs and transcripts. One thing that's very important to us is being able to get back to the actual source rather than just accepting whatever an AI summary says. For transcripts, for example, we're working around exact references so a researcher or lawyer can go back to the relevant page/line and verify what was actually said. And then there is: Trial Playground https://trial.lawnova.pro This is a different experiment. We're looking at how AI can help organize trial preparation and simulate parts of the trial workflow rather than simply generate documents. We're experimenting with things like case preparation, voir dire, arguments, evidence, witnesses and different trial scenarios. There are other pieces we're developing as well — legal research, CourtListener integration, citation verification, appellate workflows, local/private AI models and eventually more automation between all of these systems. But I'm actually not posting this because I think we've figured it all out. Quite the opposite. Things in AI are changing so quickly right now that I'm trying to figure out which parts of this are genuinely useful and which parts only look interesting because we're the people building them. So I'd really appreciate people here taking a look and telling me what you think. I'm especially curious about a few things: Does the overall idea make sense when you first see it? Is it immediately clear what each product is supposed to do? Is there something that feels unnecessary or overcomplicated? Is there something obvio

  • 情报分类：商业与市场研究
  • 分类依据：内容涉及商业、投资或市场动态
  • 信息来源：Reddit · SideProject
  • 发布时间：2026/9/21 06:44:45