Nov. 28, 2025
Blawx version 2.0.0 is now live at app.blawx.dev!
Blawx began as a demonstration of some ideas that I came up with while doing my Masters in Computational Law at the University of Alberta. Those ideas eventually turned into v0 of Blawx, which I used to do demonstrations and talks about the future of Rules as Code. Fundamentally, the idea was that someone should take the power of logic programming and improve the user interface to make it easy for non-programmers to adopt for legal purposes. Blawx was my illustration of what that might look like.
From 2021 to 2023 I had the opportunity to work on Interchange with two departments of the Government of Canada, where Blawx (v1) was re-implemented as an open source web app for use in public sector experiments led by the Canada School of Public Service.
At the end of my term with the Government of Canada LLM technology had dramatically improved, and I was interested to see whether there was a way to have large language models (LLMs) contribute to the time-consuming and challenging task of drafting Rules as Code encodings in the Blawx visual interface. Those initial experiments were so promising that I decided to re-implement Blawx as a freemium web application.
Blawx v2 is that web app.
I'm a former lawyer in Canada, working for the last several years as a legal technologist focused on generative and symbolic AI. As mentioned, I did a masters degree in Computational Law at the University of Alberta on symbolic AI, studying how to enhance the usefulness of declarative logic programming in law. I'm passionate about Rules as Code, the public sector movement toward automating reasoning over statute and regulation to improve policy design, legislative drafting, software development, service delivery, and compliance. I'm also very excited about the potential for combining modern generative AI with the symbolic artificial intelligence techniques that have been recognized as having so much promise in the legal field since the advent of logic programming in the 1980s.
Blawx is for turning rules into software. That has a lot of potential uses, but some of the big ones include:
"Rules as Code" is the specific use-case that Blawx's features and user interface are built around, but the underlying technology can be used for any situation in which it is helpful to have a computer be able to do sophisticated reasoning over a set of written rules.
So far, the people who have found the most value in Blawx are people who work in the public sector and are tasked with policy design, legislative or regulatory drafting, or developing systems that must adhere to legislation and regulation, such as benefit eligibility tools. It has also been useful for experiments in computational law, and I'm hoping that it will find greater use among legal researchers and law students who are interested in the formal representation of legal knowledge.
Blawx is specifically designed for people who aren't programmers, and don't want to be programmers. Take the example of an electronic spreadsheet. Arguably, spreadsheets are a "user-friendly declarative domain specific programming environment for math and finance." But we don't usually think of them that way. We think of them as easy-to-learn apps that are useful for math. That's how I want Blawx to feel, except for rules instead of formulas!
Not at all. Blawx is "freemium" software. The basic features are enough for learning and experimental purposes and are completely free. Advanced features, including integrations with generative AI, are available to paid subscribers.
Blawx is a passion project that I've spent the last two years slowly building evenings and weekends. It's indie, small-batch, home crafted, bootstrapped legal technology!
I'm having fun, but ultimately I want to test my hypothesis - that if you make declarative logic programming for law easy to learn and use, that legal experts will find value in it. The accepted wisdom in the tech startup world is that if you wait until you are not embarrassed to release your product, you have waited too long for feedback, and the most trustworthy feedback is revenue.
I'm focused on learning, not avoiding embarrassment. So I'm releasing it as it is, now.
Plus there are non-negligible hosting and inference costs associated with deploying it this way. So any revenue subsidizes the availability of the free subscription for the rest of the world.
Now that I've done the humble bit above, I'll do the braggy bit here. There are no tools of which I'm aware that come close to the combination of ease-of-use and sophisticated symbolic artificial intelligence offered by Blawx. Some products you might be aware of that overlap with Blawx in functionality include AustLII's DataLex, Oracle's Oracle Intelligent Advisor, Neota Logic, and programming libraries and languages such as L4, Catala, OpenFisca, Policy Engine, and others.
The main distinguishing features of Blawx to most of these other tools are these:
Now that Blawx is live, I can step away from productizing it, and step back toward improving it. There is a long list of improvements that needs to be made, and I'll try to keep this Blawg active with information about the roadmap.
In the near term, my focus is probably going to be on improving the generative AI-powered code generation features, which was the reason for re-implementing Blawx in the first place. I will also be trying to add to the beginner resources, with more example projects and tutorials.
I'm also moving into a world of maintaining and supporting a freemium indie SaaS product, which is something that I've never done before. So I expect new and interesting bumps in that part of the road, too.
Create an account, open up the beginner's guide, see what happens, and stay in touch.
Happy Blawxing.