A PRACTICING LAWYER’S WORKSHOP
I’m Pooya Shoghi, a patent attorney with a background in computer science and software development. This is where I explore how AI can support the substance of legal work.
Explore the projectsSelected projects, experiments, and work in progress.
Claims / Specification
Record / References
THE WORKBENCH
From a better disclosure
to a more explicit legal analysis.
A staged workflow for investigating claim meaning and examining how alternative constructions affect a claim chart. The design keeps the sources, competing interpretations, and mapping decisions available for review.
Explore the workflowInteractive workflow overview below.
A live analysis tool is being developed.
An invention-disclosure workflow that evaluates completeness against a configurable rubric, identifies missing information, and develops focused follow-up questions. Structured findings help a reviewer see what still needs attention.
The technical problem, inventive concept, components, operation, implementation detail, alternatives, advantages, and validation.
Evaluation results include criterion-level scores, explanations, missing information, and recommended questions. Application logic aggregates scores and applies the configured thresholds.
External project link to be added.
A separate demonstration edition of the Idea Explorer project is planned for this collection. It will use sample material and independent branding.
The demonstration will appear here when it is ready.
INSIDE THE FIRST PROJECT
Claim Construction
Proposed workflow · select a stage
01 / SOURCE RECORD
Collect the patent and relevant prosecution materials, and identify the sources that later stages can cite. Record gaps belong in the analysis from the beginning.
An indexed source set with document identifiers and locations that can be checked.
Review questionCan a reviewer locate the material behind each finding?
This is an interactive explanation of the proposed workflow. It does not run a model or produce a legal analysis.
BEHIND THE PROJECTS
I work at the intersection of patent practice and software development. The questions that interest me are practical: what should a tool examine, which distinctions must it preserve, and how can a lawyer evaluate its conclusions?
These projects are a way to work through those questions. My aim is to turn a professional method into explicit steps, test the connections between them, and make the resulting work easier to examine and improve.