AlphaClaim finds, reads, and claim charts
tens of thousands of documents automatically.

For invalidity or infringement identification.

"We found AlphaClaim generated quality output in a real-world invalidity situation."
- Dr. Jay Fu, Partner, Weaver Austin Villeneuve & Sampson LLP

By watching any of the below videos, you agree to the gateio app.

Powered by OpenAI's newest flagship model GPT-4o ("omni").
Meet us at:

Thorough (In-)Validity Determinations (35 U.S. Code §102 / §103)

Automatically claim charts thousands of references, including patents, PhD theses, scientific papers, source code, and books, leaving "no stone unturned". AlphaClaim-generated claim charts can be found in case study (single-reference), and case study (combination of 3 references).

Exhaustive infringement identification

Claim chart 1 patent or an entire portfolio against patents, product manuals, documentation, source code, blog posts, and books, to find the most likely infringements. AlphaClaim-generated claim charts can be found in case study.

Case Study Reports with AlphaClaim-generated Claim Charts

These reports were created using AlphaClaim, for illustrative purposes only, without any third-party compensation or under any client contract or direction. See our terms.

Invalidation Case Study: Patent US 7,154,905 B2

This patent has been asserted against at least Dell, HP and Lenovo, with all three cases currently in litigation. No PTAB cases have been filed yet. In about five hours, AlphaClaim automatically, accurately and exhaustively claim charted 3,000 references and was able to find single-reference invalidity ground. Download the report here (single reference).

Infringement Case Study: Patent US 8,332,844 B1

This patent is owned by Intellectual Ventures (IV) and has been asserted against at least JP Morgan, Liberty Mutual and Comerica. In about one hour, AlphaClaim claim charted 1,081 documents to look for potential infringement and found two documents that may suggest potential infringement on the part of two large multinational corporations in core product areas. These are completely distinct from the supporting technologies being pursued by IV. Download the report here.

§103 Invalidation Case Study: Patent US 10,355,975 B2

This patent has been asserted against Cerebras Systems, which claims the world's fastest AI chip. Cerebras filed an IPR; however, the PTAB's Final Written Decision found none of the challenged claims unpatentable. Litigation is ongoing as well. In about eight hours, AlphaClaim automatically, accurately and exhaustively claim charted 2,500 references and was able to find a high-scoring three reference invalidity ground. Download the report here (combination of 3 references).

Office Action Case Study: Application 18/500,070 (US 20240061713A1)

This application was filed by Microsoft/Fungible on 11/01/2023 and published on 02/22/2024, without any Office Action so far. In about one hour, AlphaClaim automatically, accurately and exhaustively claim charted 923 references and was able to find a high-scoring two reference invalidity ground. gate io login

FAQ

What can AlphaClaim do?

AlphaClaim's capabilities are valuable in several scenarios:

  1. To identify the best invalidity grounds (including combinations), for IPRs and Office Actions. For litigated examples with AlphaClaim-generated claim charts, see Case Study Report for US 7,154,905 B2 and case study report for US 10,355,975 B2. For an Office Action example with an AlphaClaim-generated claim chart, see Case Study Report for Application 18/500,070 (US20240061713A1).
    ‍
  2. To identify evidence of patent validity prior to assertion or sale of a patent. If AlphaClaim’s exhaustive review turns up low AlphaScores for all documents, a patent owner can be more confident that the patent will survive IPR or other 102/103 challenges.
    ‍
  3. To identify the best evidence of potential infringement, for patent owners. See Case Study Report for US 8,332,844 B1 for an example with an AlphaClaim-generated claim chart.

How does AlphaClaim work?

AlphaClaim's patent-pending AI technology uses multiple AI models, statistical sampling, context extraction and multidimensional scoring. More information about other similar AI systems, such as Google's AlphaCode2 and AlphaGeometry, can be found in this post from the Berkeley AI Lab.

How accurate is AlphaClaim?

AlphaClaim's patent-pending AI architecture has been aligned and optimized to match the quality of skilled human experts. In particular, an AlphaScore of 4 or 5 corresponds to evidence stronger than the median IPR petition (for invalidity) or the median filed claim chart (for infringement).

Is AlphaClaim really 100% automated?

AlphaClaim relies on IP attorney input for claim construction, which can be in the form of suggestions or ideas or more formal. For a typical patent, it takes about 15 minutes to enter claim construction specifics into AlphaClaim. After that, AlphaClaim can find, read, and claim chart 10,000s of references completely autonomously and return the top scoring claim charts.

AlphaClaim can also help explore the impact of various claim constructions prior to a Markman hearing.

Is AlphaClaim fast?

AlphaClaim makes it practical and realistic to assess patent invalidity, validity and infringement with pinpoint accuracy. AlphaClaim enables 10,000s of documents to be evaluated and claim charted within hours at reasonable cost. In the past, such a "no stone unturned" analysis of a patent would have cost millions of dollars, per IPWatchdog's Gene Quinn.

Does AlphaClaim hallucinate like ChatGPT?

No. AlphaClaim's patent pending technology completely eliminates hallucinations that may affect claim chart excerpts.

Who are the inventors of AlphaClaim?

AlphaClaim was developed by PriceWire, Inc., a team of AI researchers formerly with the Berkeley AI Research Lab (BAIR) and Meta.