Patent landscape analysis with AI
Map a technology field from a plain description: technology clusters, key players, filing trends, geography and a citation network, with a written report.
A technology landscape maps a whole field of patents rather than ranking documents against one patent. You describe a technology in plain language, and One Transform gathers the patents in that area, groups them into technology clusters, ranks the companies filing in it, charts filing activity over time and across countries, and writes a briefing on the field. Strategy teams, R&D leads, IP counsel and analysts use it to understand a field before investing, filing or entering it.
What you get
- An overview: patents mapped, number of technology clusters, unique assignees, a maturity reading for the field and how concentrated ownership is, with an executive summary.
- Key players: assignees ranked by the number of patents they hold in the field, with whether their filing is growing, flat or declining.
- Filing trends over time, including actual US filing volumes by year from USPTO data, so the chart reflects the field rather than only the sample the search collected.
- Technology clusters, named after the technology they contain, each with its patents, its top assignees and a written analysis.
- Geography: where the field is filed, by jurisdiction.
- A citation network showing how the key patents relate.
- A full report covering competitive dynamics, technology maturity, filing trends, geographic filing strategy, emerging and declining sub-areas, standards and standard-essential patents where relevant, and recommendations.
- The patent list with a relevance score for each patent and its breakdown.
- Exports to Excel, PDF and Word with the same data.
How it works
- Scope. Your description is turned into the field's vocabulary, likely classifications and sub-technologies. You can anchor the map on a seed patent, and add keywords, assignees, classifications, jurisdictions, a date range and patent status.
- Broad discovery. The search casts a wide net on purpose: meaning-based search, classification and keyword searches, inventor trails, a decomposition of the technology into parts, cross-domain applications, searches across time periods, academic literature, citation and co-citation links, standards-related filings and recent US grants from USPTO data.
- Classification data. Patents found through search often lack their classification codes, so codes, citations and family data are fetched for the scoring pool before it is scored.
- Relevance scoring. Each patent is scored for relevance to the field. Classification overlap, citation proximity and commercial signals are computed from that patent data; technical similarity and timing are model judgments.
- Analytics. Clusters are formed by grouping patents on the meaning of their titles and abstracts and then named, which avoids clusters that are just a classification code. Assignee rankings, trends, geography and the citation network are computed from the patents.
- Report. The executive summary, cluster and assignee analyses and the full report are written from the computed analytics.
When to use it
- Before an R&D investment, to see how crowded a field is and who leads it.
- Before entering a market, to see which companies and countries matter.
- For portfolio strategy, to see where your filings sit against the field's clusters.
- For a board or investor briefing that needs a defensible picture of a technology area.
A landscape describes a field. To profile one company's position inside a field, run competitive intelligence. To find the thinly filed parts of a field, run a white space analysis. To start from a single patent and find its neighbours, use similar patents.
Limits
- The map is built from a large sample of the field, not a census. Cluster sizes and assignee rankings are shares of the patents found; the US volume figures come from USPTO counts.
- Applications publish about 18 months after filing, so the last two to three years of any trend chart are incomplete. Why recent patent filing trends look like a decline explains how to read them.
- Assignee names are not always consolidated across subsidiaries and acquisitions. Check the key players list before quoting market shares.
- The briefing is a research summary. It is not legal advice or an investment recommendation.
Frequently asked questions
Do I need a patent number to map a field?
No. A plain-language description of the technology is enough. A seed patent is optional and anchors the map around one invention.
How are technology clusters formed?
Patents are grouped by the meaning of their titles and abstracts rather than by classification codes alone, and each group is named after the technology it contains. You can also choose to group by assignee, by date or by citation network.
Why do the filing trends show US volumes separately?
Search results are a sample, and samples undercount recent years. Actual US filing volumes by year come from USPTO data and are shown alongside the sample, so you can tell a real slowdown from a sampling effect. How to read a patent landscape report covers the other checks to make.
Can I restrict the landscape to certain countries or years?
Yes. Set jurisdictions, a date range and patent status before the search runs. Assignees can be included or excluded as well.
How much does a landscape cost?
A landscape counts as one search on your plan, like every other search type. See pricing.

