Skip to content
View Varietyz's full-sized avatar

Block or report Varietyz

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Varietyz/README.md

Bane's Lab

Bane's Lab

Systems architect, technical lead and R&D · Jay Baleine - Ostend, Belgium


I design and govern software systems: greenfield builds, legacy modernization, architectural refactoring and automation, in whatever language or domain the system calls for. Much of that code is now written by an LLM, so the other half of my work is the governance around it. I build automated checks that hold the architecture in place, keep complexity controlled and dependencies minimal, and context architectures that give a model structured project context as the system grows.

I have been self-taught since 2012. I started in graphic design and moved through hardware, systems and software development to research on how to work with models.

Work

  • Bane's Lab · System Architect. My own practice, a software and AI consultancy for the governance, design, development and transformation of software systems.
  • Kinsugi.ai · Co-Founder. Platform architecture, ontology and knowledge engineering, and agent governance.
  • Beasr · Technical Lead. A relocation platform that brings the decisions and purchases of a home move into one place.

Projects

A method for building software with an LLM that writes most of the code. Its rules are held by automated checks instead of reminders, the project's state lives in files instead of chat history, and a claim about the code counts only once the developer or the model has read the file. It runs through six chapters: Start, Plan, Build, Verify, Collaborate and Ship. banes-lab.com serves every page as HTML, Markdown and JSON, so a model can fetch the chapter it needs while you work.

A structured format for writing instructions to LLMs, defined by a formal grammar that is grounded in a reasoning ontology and published with a guide and a set of templates. I use it for agents, planning templates and validation gates, the parts of the work that have to run the same way every time.

Focus

  • Governed platform architecture
  • Knowledge ontologies and reasoning systems
  • AI agent governance
  • Polyglot quality infrastructure
  • Full-stack SaaS engineering
  • End-to-end DevOps
  • Application security and offensive testing
  • Documentation as code
  • Toolchain automation

Methodology PAG Architecture Ontology Anatomy FAQ

Certifications

AI & Prompt Engineering

Machine Learning & Development

Activity

GitHub activity: contributions, commits, pull requests, stars, repositories, followers, languages and weekly contributions

Contact

Website Email LinkedIn Discord

Pinned Loading

  1. Disciplined-AI-Software-Development Disciplined-AI-Software-Development Public

    Disciplined AI Collaboration: a method for building software with AI. One loop at every size, rules held by checks rather than attention, state derived rather than written, evidence in place of cla…

    HTML 409 34

  2. discord-js-bot-template discord-js-bot-template Public template

    Scalable Discord Bot Template, includes: GitHub Workflow (Actions + Releases), API, Rate-limiting, Regression detection, benchmarking/stress-testing, Hot reloading, Plugin architecture and rigorous…

    JavaScript 17 5