Crystal aOS

A best-practice framework for AI adoption and governance in regulated firms

‍

‍
‍

AI governance for law firms and financial services firms

Crystal aOS enables legal and compliance teams to use AI responsibly. Use our best-practice resources to build or uplift AI governance in your firm.
‍

Get in touch

What we do

We teach and implement our AI framework that is purpose-built to align with international data security and AI safety standards like ISO 27001 and ISO 42001.

Framework

An end-to-end reference checklist for best-practice AI use in regulated environments.

Resources

A set of AI policies, SOPs, templates and registers for governed AI operations.

Education

AI skills training sessions for board, executive and operational staff, off-the-shelf or tailored for your firm.

Implementation

Practical AI adoption assistance: PTEs and FTEs in your firm to establish, monitor, and maintain the safety and compliance of your AI system.

The problem

Chief Legal Officer / Chief Compliance Officer

“I need to use AI to uplift productivity in our legal and compliance work but my team don't have the knowledge or time to build and maintain AI safety, data security, privacy and governance controls.

We have no visibility of the risks being introduced by the AI-coded apps we've already built internally.”

Take the next step

Get the framework to run yourself, or have us run it with you over an initial 90-day period.
‍

Get in touch

Why we exist

To simplify legal and compliance so that responsible innovation can thrive.

Joni Pirovich

Founder

In the last decade I founded APAC's leading blockchain and digital assets specialist law firm, have acted for pioneering emerging tech platforms, and even foreign governments.

I've seen first-hand how difficult it is for legal and compliance operators in emerging tech (and traditional businesses) to apply existing laws at speed, let alone keep up with new tech-specific ones popping up around the world.

My first AI build was Crystal Chat - a generative AI chat tool that spoke with my repository of global crypto-specific laws, cases, and reg guidance materials. It amplified me as a lawyer but after 6 months of super charging ~20 hours of work into an 8-10hr human work day I realised humans driving genAI based tools was not sustainable.

So I went back to the drawing board and designed a system that could "containerize" the knowledge, skills, experience and processes required to do regulated work without me having to constantly be "at the wheel". These containers - I call them Agentic Assets - are the necessary building blocks to realise the promise of secure AI use with operational goverance.

Like the tech for a self-driving car must mature it's level of autonomy through oversight, evaluation and continuous improvement, so too must Agentic Assets go through an intentional process and secure system. This is the foundation built in Crystal aOS as a best-practice reference framework for AI adoption in regulated firms.