The Challenge That Started It All

Every patent attorney knows the process: manually reading through hundreds of patent claims, weighing each word, analyzing every limitation, trying to determine which patents are truly valuable.

It's expertise that can't be rushed, can't be automated—or so everyone thought.

Until David Andrews decided to change the game.

ClaimScore emerged from a simple but powerful idea: what if we could teach machines to evaluate patent claims the same way experienced attorneys do, but at massive scale?

Born from Necessity, Proven at Scale

Developed by David Andrews—a unique blend of software engineer and patent attorney—ClaimScore represents the perfect fusion of deep technical expertise and sophisticated legal understanding.

David's background spans a decade at Microsoft contributing to Windows kernel development and Age of Empires, followed by summa cum laude law school graduation and judicial clerkship.

ClaimScore uses patent-specific natural language processing to replicate the nuanced analysis that attorneys perform by hand when evaluating claim breadth.

Instead of reading one patent at a time, ClaimScore can analyze thousands—identifying the strongest claims, the broadest coverage, and the hidden gems that others miss.

NLTK-Powered Engine Optimized for patent language
Custom Grammar Rules Patent-specific structures
ML Models Trained on litigation outcomes
Scalable Architecture Thousands of patents efficiently

The Uber Validation

The first major test came when Uber faced a critical challenge: preparing for their IPO while building a defensive patent portfolio in a competitive market.

The Problem Was Massive:

  • Hundreds of patent portfolios being offered for sale
  • Thousands of individual patents to evaluate
  • Limited IP team resources
  • Need to focus on quality over quantity
  • Pressure to find assets not yet on the broader market

ClaimScore Delivered Results:

Working directly with David Andrews, Uber integrated ClaimScore into their acquisition process.

Prior to manual review of any patents, we run each offered portfolio through analysis software that implements our defined methodology to identify relevant families and the best claim from the highest-ranked patent in each family.
— Kurt Brasch, Uber Patent Transactions Head

The outcome? Uber successfully built a strategic patent portfolio through "savvy deal-making," finding quality assets while others were still sorting through the noise.

How ClaimScore Works

A sophisticated four-step process that transforms patent claims into actionable intelligence

1

Intelligent Element Recognition

Automatically identifies distinct claim elements—the building blocks that matter in infringement analysis. Distinguishes between new limitations and references to previously cited elements.

2

Linguistic Structure Analysis

Creates parse trees showing how claim elements relate to each other. Maps the logical flow of claim limitations and identifies critical dependency relationships.

3

Weighted Scoring Algorithm

Assigns numerical scores based on linguistic characteristics. Applies attorney-knowledge weightings for restrictive language. Lower scores indicate broader, stronger claims.

4

Benchmarked Validation

Compares scores against database of litigated patents. Correlates claim strength with actual trial outcomes and provides probability estimates for infringement findings.

Applications Beyond Patent Acquisition

ClaimScore's versatility extends across the entire patent lifecycle

Portfolio Management

  • Identify your strongest patents across entire portfolios
  • Compare claim strength within patent families
  • Prioritize prosecution and maintenance decisions

Competitive Intelligence

  • Analyze competitor patent strength at scale
  • Identify market gaps and opportunities
  • Track portfolio evolution over time

Litigation Support

  • Quickly identify strongest claims for assertion
  • Assess infringement probability before filing
  • Support claim selection and case strategy

Technology Assessment

  • Evaluate patent landscapes for M&A due diligence
  • Assess freedom to operate across technology areas
  • Benchmark patent quality across different sectors

Ready to Scale Your Patent Intelligence?

Whether you're building a portfolio, analyzing the competition, or preparing for litigation, ClaimScore gives you the same analytical power that Uber used to build their billion-dollar patent strategy—but applied to your specific needs.

David Andrews now brings this proven technology to clients through Novelty Hill IP, the firm he co-founded with Jack Bradley in 2024 to answer complex IP questions using the latest in machine learning and AI combined with decades of IP experience.

Contact Novelty Hill IP

About the Inventor

David Andrews is a distinguished patent attorney and machine learning expert who co-founded Novelty Hill IP in 2024. His unique background combines a decade of software engineering at Microsoft (contributing to Windows kernel and Age of Empires) with extensive legal expertise gained through summa cum laude law school graduation, judicial clerkship, and high-stakes patent litigation practice. David holds multiple patents including the ClaimScore Algorithm and continues to innovate at the intersection of AI and intellectual property law, frequently speaking at conferences and participating in technology hackathons.