Legal modelling
I structure legal rules, conditions, exceptions and dependencies into a clear path of reasoning.
Legal Engineer
LAW. LOGIC. TECHNOLOGY.
I connect legal reasoning with technology to build predictable AI agents that support legal professionals. With explicit rules, controlled actions and results that can be checked.
01 / PURPOSE
The goal is to develop agents that support legal professionals by combining AI models with deterministic decision logic. The model helps interpret and organise information. Explicit rules define the next steps, the conditions for taking them and the points that require a lawyer’s decision.
Predictability rests on rules: the same approved inputs and rules should produce the same result in the rule-based layer. The workflow can be traced, uncertainty stays visible and the lawyer retains control over the decision.
02 / WORK IN PRACTICE
Legal reasoning, made explicit.
I develop Subsumio — a workspace for legal reasoning paths and subsumption. I turn legal rules into structured conditions and relationships, and facts into an analysis with reasoning and a decision history.
This provides a foundation for agents whose actions should follow defined legal reasoning paths and remain open to verification.
Conditions, exceptions and logical operators in a clear reasoning model.
Yes, no or unknown. Missing information remains an explicit part of the analysis.
Snapshots, decision history and access controls for matters.
03 / APPROACH
I structure legal rules, conditions, exceptions and dependencies into a clear path of reasoning.
I design workflows and checkpoints that connect AI models with explicit rules and the lawyer’s tools.
I design for explanations, versions and change history. A result should be reproducible and open to review.
04 / CONTACT
For AI agents that support legal professionals, legal tech products and collaboration on predictable legal tools.
Send a message