LegalQuants
Leonard Hope

Leonard Hope

U.S. Patent Attorney

Using AI to Build Tools for Patent Attorneys

About

I'm a U.S. patent attorney with a background in computer engineering. Over the past several months, I've been using Claude Code to build tools for patent work: USPTO and BigQuery search skills, an examiner intelligence dashboard, a patent landscape report generator, a filing-package QC tool, and an AI assistant that joins patent disclosure meetings to help with real-time question coverage. What drives the work is the same idea behind all of it: patent practice is full of small, fixable inefficiencies that compound across a portfolio, and the tools to address them are finally within reach for individual practitioners.

7 Projects

Claude Skill for Patent Filing Quality Control

Claude SkillOpen SourceSource

A comprehensive quality control tool for U.S. patent application filing documents. This Claude Code skill systematically checks for internal document errors, cross-document inconsistencies, USPTO compliance issues, and common filing mistakes before submission to the USPTO.

USPTO ODP Natural Language Search

Claude SkillOpen SourceSource

A Claude Code skill that provides natural-language access to all USPTO Open Data Portal APIs for patent and trademark data. Ask questions in plain English -- the skill handles API selection, query construction, and result formatting automatically.

Google Patents Natural Language Search

Claude SkillOpen SourceSource

A Claude skill for searching Google Patents using plain English. Ask questions in natural language and get results from 166M+ patent publications across 17+ countries — powered by BigQuery under the hood.

Claude Skill for Patent Landscape Analysis

Claude SkillOpen SourceSource

A Claude Code skill that produces decision-grade patent landscape reports as self-contained interactive HTML files. Describe a technology area in natural language — the skill pulls the relevant patent data from Google Patents BigQuery, runs the analytics, and writes a single .html file you can share with stakeholders. The file opens offline, stays interactive forever, and every number in it has receipts.

Pat — AI Patent Disclosure Meeting Assistant

Web AppOpen SourceSource

Pat is a real-time AI assistant for patent inventor disclosure meetings. Pat listens to the meeting, transcribes on-device, and provides live suggestions telling the patent attorney what questions to ask — before the conversation moves on.

Taskspect

Desktop App

A task management app with three synchronized views that give you different perspectives on the same tasks: Task View — column-based project layout Timeline View — Gantt-style timeline Timebox View — day planner with hourly time slots Press Tab to cycle between all three views.

Claude Skill for Patent Examiner Intelligence

Claude SKillOpen SourceSource

Patent Examiner Intelligence is a Claude skill — an AI-powered tool that a patent attorney invokes through natural conversation with Claude. The attorney asks Claude to profile an examiner, and the skill handles the rest: fetching data from the USPTO's public APIs, computing the analysis, and producing an interactive HTML dashboard that can be opened in any browser, emailed to a colleague, or archived with the case file.

Philosophy

"AI Can Help With Efficiency, But the Real Opportunity is Improving Quality"

Most of the AI conversation in legal has been about efficiency: faster drafting, faster review. That's real, but the bigger opportunity in patent work is quality. A patent is a long-lived asset whose value is set at specific moments: the disclosure, the claim scoping, the prior art positioning, and the response to a particular examiner. AI tools, used well, let individual practitioners invest more attention at those points, not less. The goal is not to do the same job in half the time. It is to deliver a meaningfully better patent. That is where the real client benefit lives, and what I am building toward.