The platform integrates invention parsing, multi-LLMs, parallel web searches, and a verification layer to identify potential infringing products for any patent.
MOHALI, PUNJAB, INDIA, May 7, 2026 /EINPresswire.com/ — ClaimHit, an AI-driven platform for patent infringement searches and claim mapping, has announced the release of v2 of its analysis engine. This patent-pending tool is designed for patent attorneys, IP licensing teams, and in-house IP counsel who need to rapidly identify products that might infringe on a patent’s claims. Traditionally, this process has required days or weeks of manual effort.
The public release of v2 comes as the core pipeline now delivers results that the team is confident in sharing. Furthermore, the practitioners who perform this work daily are the best judges of what features should be developed next.
“V2 represents our current platform, but it isn’t the final goal,” said Bikram, Founder of ClaimHit. “Over the past few months, we’ve focused on the aspects we believed we could develop first: extracting inventive contributions from patents, running multiple frontier models simultaneously during candidate discovery, and verifying each candidate against the manufacturer’s data before presenting it to a user. There are still many features we haven’t implemented. We want feedback from patent practitioners on what we should prioritise, what we’re missing, and if our tool integrates well into real workflows. Our future plans will be shaped by these discussions, not assumptions on our part.”
What’s working in v2:
Invention parsing: ClaimHit automatically extracts the inventive contribution and key claim elements from a patent, eliminating the need for users to pre-process the patent before scanning.
Single-patent and portfolio scans: Supports detailed analysis of a single patent or an entire portfolio against the platform.
Multi-model LLM retrieval: Multiple independent frontier models run candidate discovery simultaneously, enhancing recall.
Parallel keyword and semantic web search: Executes lexical and meaning-based searches together across the open web.
Evidence verification before display: Each candidate is independently checked against manufacturer-published product info before being ranked or displayed.
Evidence-based ranking, with Good Match and Possible Match tiers: Candidates are ranked based on the strength of supporting evidence, not ensemble votes. Good Match indicates strong evidence; Possible Match suggests candidates worth further review with weaker evidence.
AI Hit Charts: Provides structured claim-element-to-product-feature mappings with citations, available on demand.
Sub-two-minute scans: The majority of full scans are complete in under 120 seconds.
What we’re still working on:
ClaimHit is currently focused on patent infringement searches and claim mapping. They plan to add real-time progressive results, more detailed portfolio analytics, expanded verification of all potential matches the system finds, and various workflow integrations. However, these features are not yet available. The team is cautious about promising too much on the roadmap, preferring to wait until they receive user feedback on what features would be truly helpful.
Where we’re heading:
Patent infringement search was the initial problem ClaimHit aimed to address, but the team’s broader long-term goal goes beyond that. Many patents including newly issued ones, deep-tech patents from research institutions, and patents on emerging technologies, currently have no infringers because the technology is too early-stage or the commercial ecosystem is not yet developed. For these patents, more relevant questions are future-oriented: which companies could benefit from the technology, which nearby startups might license or acquire it, which new industries could emerge, which tech-transfer partners would be ideal, and which industry experts or boards are already considering this technology.
The ClaimHit architecture in v2 featuring invention parsing, multi-model retrieval, and evidence verification serves as a crucial foundation for broader work, but it’s not enough on its own. Determining “who could commercialize this” or “who would be the right tech-transfer partner” requires additional layers that the team hasn’t yet developed: assessing companies’ technical capabilities, reasoning about adjacency and commercial readiness, and evaluating strategic fit. The team views v2 as the ground floor of a much taller structure. None of the upper floors are completed yet. They represent future development goals, and a key reason for opening v2 to feedback now is to involve those who do this strategic work daily- their insights should guide what is built next.
Pricing and how to try it
ClaimHit v2 is available today at [claimhit.com]. New users get three credits free on signup. During the v2 introductory period, additional credits are priced at $99 each. The free credits are intended to give practitioners enough to try the platform on their own patents.
ClaimHit’s evidence-first claim-mapping methodology is patent pending.
The team is actively gathering feedback from practitioners and is open to conversations with anyone who wants to weigh in on direction as a user, an advisor, a workflow partner, or a critic. If you’ve worked on infringement search, licensing target identification, freedom-to-operate, portfolio monetisation, tech transfer, or commercialisation strategy, and have views on what tools like this should do, please get in touch. Channels are linked at ClaimHit.
Bikramjit Singh
Scintillation Research
+91 9888996262
email us here
Visit us on social media:
LinkedIn
YouTube
X
Legal Disclaimer:
EIN Presswire provides this news content “as is” without warranty of any kind. We do not accept any responsibility or liability
for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this
article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
Media gallery
