Software · AI · SaaS
Know what already exists before you invest in what comes next.
Our prior art search services help technology companies identify relevant patents, technical publications and earlier approaches related to a software or AI invention — so patent and R&D decisions are made with the existing knowledge landscape in view.
Industry challenge
01
02
03
04
Scope of work
A
Searches of relevant patent publications and available patent records across jurisdictions and databases suited to the invention and its market.
B
Research papers, conference proceedings, standards documents, technical documentation and other public technical sources appropriate to the technology.
C
The invention is broken into meaningful technical features — data flows, processing steps, system components — and compared against the references found.
D
For each key reference: why it may be relevant, which technical features overlap and where meaningful distinctions may exist.
E
An organised report with bibliographic details, source references, publication dates where available, relevance summaries and feature-level comparisons.
F
A walkthrough of the findings and possible next steps: refining the disclosure, discussing claim scope with patent counsel, or a focused follow-on search.
India
Section 3(k) of the Indian Patents Act, 1970 excludes mathematical methods, business methods, computer programmes per se and algorithms from patentability. It does not make all software unpatentable. Analysis of a computer-related invention looks at the substance of what is claimed — not the product description, marketing language or labels such as "AI-powered".
Section 3(k) assessment and prior art analysis are related but distinct. A search can inform novelty and inventive-step discussions; it does not decide whether claimed subject matter is excluded. Adding hardware does not automatically overcome Section 3(k), and a technical effect is not an automatic guarantee of patentability. Novelty, inventive step, industrial applicability, excluded subject matter and disclosure requirements are assessed as separate but interacting considerations.
Official sources: Patents Act, 1970 (IP India) · IP India examination guidelines, including CRI guidelines. Always rely on the current version published by the Office.
A SaaS company develops an AI-based system that reduces latency in distributed inference by changing how workloads are scheduled across computing resources.
A useful search investigates earlier scheduling techniques, distributed inference architectures, resource allocation mechanisms and related publications. The analysis compares the disclosed features with the references found, and the Indian patentability requirements — including Section 3(k) — are considered separately.
Contrast this with a claim that merely uses a generic computer to perform an abstract business process or mathematical operation, without establishing a qualifying technical contribution.
United States
Two US Supreme Court decisions — Mayo Collaborative Services v. Prometheus Laboratories, Inc. (2012) and Alice Corp. v. CLS Bank International (2014) — form part of the background to patent eligibility under 35 U.S.C. § 101. Eligibility depends on the actual claim language and the applicable legal authorities.
A company develops a machine-learning pipeline that improves distributed data processing through a specific technical mechanism. The search investigates prior implementations, comparable data-processing architectures, model execution techniques and published technical improvements, and maps them to the invention's features. Eligibility under § 101 remains a separate legal assessment.
Alice/Mayo is a US legal framework; Section 3(k) is an Indian statutory exclusion. They are not equivalent tests and are not interchangeable.
References: Alice v. CLS Bank, 573 U.S. 208 · Mayo v. Prometheus, 566 U.S. 66 · USPTO MPEP § 2106 (eligibility)
Workflow
01
Review the technical disclosure, the problem it solves, how it is implemented and what appears to distinguish it.
02
Identify terminology, synonyms used across research and industry, related technical fields and suitable jurisdictions.
03
Investigate patent publications and relevant non-patent literature, including papers and technical documentation.
04
Compare important features against relevant disclosures and identify potential overlaps or gaps.
05
Provide a structured report and discuss potential next steps with your technical and IP teams.
Deliverables
Who it's for
Check whether a model pipeline, retrieval method or inference optimisation has already been described before committing filing budget.
Understand earlier platform architectures before positioning a product feature as a protectable differentiator.
Find prior orchestration, scaling and tooling approaches that may shape how an invention should be framed.
Review earlier detection, encryption and data-handling disclosures before investing in a patent programme.
Give founders and investors a clearer view of the existing technical landscape ahead of fundraising or filing.
Screen invention disclosures early so R&D and filing effort focuses on the most distinctive work.
Receive organised, feature-mapped evidence that supports drafting and claim discussions with the client.
Start a search
Please don't share confidential source code, unpublished invention details or trade secrets when booking. A short, non-confidential summary is enough to scope the work; detailed disclosures follow once an engagement and confidentiality terms are in place. See our privacy policy.
FAQ
A search for a software invention typically covers patent publications plus non-patent sources such as research papers, conference proceedings, standards and technical documentation. The findings are mapped to the invention's technical features — architecture, data flows, processing steps — rather than to product marketing terms.
AI concepts are described with inconsistent vocabulary across patents, academic papers and industry writing. A useful search combines keywords with classification codes, concept-based and semantic retrieval, and citation review, and then relies on an analyst to judge which disclosures actually describe a comparable technical approach.
Section 3(k) of the Patents Act, 1970 excludes mathematical methods, business methods, computer programmes per se and algorithms. It does not mean all software is unpatentable. Whether a computer-related invention is excluded depends on the substance of what is claimed, assessed under the Act and the current official CRI guidelines. Adding hardware does not automatically overcome the exclusion.
Prior art searching asks what has already been disclosed and how it compares with the invention, which informs novelty and inventive step. Section 3(k) asks whether the claimed subject matter falls within a statutory exclusion. They are related but distinct questions; a search can inform the discussion but does not determine Section 3(k) compliance.
Mayo (2012) and Alice (2014) set out the US Supreme Court framework under 35 U.S.C. § 101: first ask whether a claim is directed to a judicial exception such as an abstract idea, then whether its elements add an inventive concept that amounts to significantly more. Generic computer implementation of an abstract idea may be insufficient. This is a US framework and is not equivalent to India's Section 3(k).
No. A search provides evidence about what has been disclosed. Patentability depends on novelty, inventive step or non-obviousness, industrial applicability, excluded subject matter and disclosure requirements, assessed by qualified patent professionals. No search can guarantee that every relevant reference has been found.
Yes. For AI and software inventions, relevant earlier work is often published in conference papers, preprints, standards or technical documentation before or instead of patents, so non-patent literature is included where it is appropriate to the invention.
Common points are before drafting a patent application, before major R&D investment in a new technical direction, before fundraising or due diligence, and when refining an invention disclosure after an earlier search or office action.
Knowing the closest disclosures helps inventors and counsel identify what appears genuinely distinctive, describe the technical problem and solution more precisely, and discuss claim scope with that context in mind. Drafting decisions remain with the client's patent counsel.
For scoping, a short non-confidential summary is enough: the technical field, the problem addressed, the general approach and the jurisdictions of interest. Detailed technical disclosures are shared only after an engagement and confidentiality arrangement are agreed.