Isaacus – the legal AI research company
Specialized legal models topping retrieval benchmarks when general LLMs hallucinate.

Searchable index for 3.65M private companies when Crunchbase charges for this data.
Investors, journalists, and researchers investigating private companies.
Crunchbase · PrivCo · OpenCorporates
I also already know the data itself is already public so how am I exepcting users, the point is the aggregation and linking everything back to the government source. What’s currently on the website is only a read-only slice of what I have running privately. It also computes things like risk tiers, financial trends such as operating-margin trajectories, watchlists/event streams, and as-of views showing what was knowable at a specific point in time. Some of those parts are still early and incomplete due to the project’s immaturity, and I intentionally left them out of the public demo for now due to posting "risk" tiers/labels on companies is kind of complicated as well. I might consider adding them to the existing demo companies if it is safe for me to do so though.
And i'm trying to scale it beyond that though, if I get any positive signs or in general anyone intersted in my project, or I simply shut it down and I might open-source the code. Would appreicate guidance/feedback, Also, if there is a company that is missing, drop It's name, I'll search it on my end and reply to you with it's record.
Specialized legal models topping retrieval benchmarks when general LLMs hallucinate.
Data-first move I appreciate: EDGAR-sourced, normalized financials plus 13F aggregation and a 'cluster buying' insider flag give immediate, actionable entry points for dividend and value investors. The UI is tidy and scannable — lists, recent updates and investor portfolios are surfaced clearly — but there’s nothing radically new here (no backtesting, unclear export/API options), so it’s a very useful tool for DIY research rather than a replacement for paid terminals.
Legal framework for AI-run companies with 36 packages for Argentine compliance.
Yet another AI job board competing with AI Jobs and Otta.
Scrapes 25 websites to aggregate water data, but solves a problem few people prioritize.
Clean API for Indian corporate data when scraping MCA usually means CAPTCHA hell.