sequenced.ai
Articles/Search & research/Blueprint//8 min read

Patsnap follows an engineering question into the patent record

Patsnap connects engineering questions to patent and scientific evidence. A latch design puts its AI search, credit offers and review process into context.

By Sequenced deskAI-assisted, source-led · how we work
Visit Patsnap website ↗
HiroResearch assistantAccepts a question, document or image.
SkillsSaved instructionsReuse a method across research tasks.
EurekaAgent workspaceSpecialist agents support IP and R&D work.
APIsData accessEmbed innovation records in an existing application.
Patsnap mark
Patsnappatsnap.com · independent research

Represent this company? Verify your work email to access its workspace, or send the desk a factual correction.

A design engineer brings a drawing of a latch to the patent team. Searching for “latch” turns up inventions whose only resemblance is the name. Searching for the product it belongs to misses a similar mechanism used in another industry. Patsnap gives that conversation a different starting point. Its Eureka software can work from a technical description, look across patent and scientific records and return references a researcher can inspect. The interesting question is whether the search changes what the engineer and patent professional know enough to justify their next hour of work.

In brief
  1. 01The company Patsnap supplies innovation data, patent analytics and AI-assisted research for intellectual property and engineering teams.
  2. 02The product Eureka's Hiro assistant can take an initial question and carry its context into a more specialized research task.
  3. 03The test Give the team one known mechanism, one awkward synonym and a previously identified reference. Inspect what the search retrieves and which features each document actually discloses.

01 / ProductWhat sits behind a Eureka answer?

Patsnap’s product portfolio spans patent analytics, specialist scientific data and the Eureka family of AI agents. An IP analyst who already knows the publication number can go straight to a record. Eureka is more interesting at the uncertain beginning of a project. A researcher may know what a part does and how it fits into an assembly, yet have no idea how earlier inventors described it. Turning a drawing and a few sentences into promising lines of inquiry is a different task from looking up a number.

The Hiro user guide describes a natural-language entry point. Users may attach a document or image, specify a scope and ask for a particular output. Hiro offers references and can suggest an advanced capability after the question becomes clearer. Skills supply instructions that can be reused across tasks. A team could specify the same columns for every mechanism comparison while letting the actual question, evidence and conclusion vary from one investigation to the next.

Patsnap also describes ways to access its underlying innovation data from enterprise applications. The hosted research interface and an embedded data service involve different work. In Eureka, the researcher can refine a question, open references and collect material for discussion. In a custom engineering portal someone must decide how those records will be displayed, which updates matter and who owns the relationship between a patent record and the company’s own component catalog.

02 / AudienceWhen an engineering team needs a patent researcher beside it

Imagine a design review scheduled for Friday. Two engineers propose different ways to release a cover with one hand. The IP specialist wants to see what each mechanism shares with prior designs before committing to a broader search. One colleague understands the spring and load path, another understands how claims describe arrangements of parts. Giving both people a common set of exhibits can keep a meeting focused on the disputed feature instead of the words that happened to appear in the first query.

A second audience is an information specialist who receives several versions of the same request. “Find competing products” might mean identify assignees, retrieve similar drawings, or locate recent publications on a manufacturing method. Those are different deliverables. Ask the engineer which decision the search should inform, whether drawings or claim text matter most and what dates belong in scope. A precise brief makes a short evidence list more useful than a long export of superficially related titles.

03 / WorkflowFollow the motion of a latch through the source records

For a proposed evaluation, use a public drawing or an internally cleared description of the latch. Identify the features the team considers essential, such as the relationship between the cover and catch and the sequence of motion during release. Keep one known reference out of the initial prompt as a retrieval check. Supply the permitted date range and ask Hiro for a comparison by mechanism, with a reference for each row. A suggested term is worth keeping if it opens a second line of inquiry even when the first list contains few useful documents.

Read the drawings and relevant passages for the nearest matches. A title about access panels may conceal a different release mechanism, while an unfamiliar industrial enclosure may show the exact joint. For each retained item, write down the publication identifier, the passage or figure supporting the match and the feature the researcher could not find. If a row relies on two separate publications, keep those observations distinct. Combining isolated features into a convincing paragraph can conceal which document actually describes the arrangement.

Ask a second researcher to revisit the disputed rows without the machine summary. Record whether the source opens, whether the figure supports the description and what the dates mean for the question being asked. A later family member can be an excellent illustration of a mechanism and still be unsuitable for a search with an earlier cutoff. The Hiro documentation describes inspecting patent details, family data and sources. The team still has to choose the relevant member and explain its inclusion.

Bring the engineers back only after that pass. A useful deliverable is a short selection of references with annotated uncertainties, alternative terminology and one design question to carry into Friday. A professional can decide whether to commission an expanded search or obtain an opinion. Keep search notes alongside the original product drawing so a subsequent revision to the latch does not quietly turn the first report into an answer about a different design.

04 / PricingTwo public price presentations use different credits

Published routeCommercial presentationScope to verify
Eureka main offerFourteen-day trial with 10,000 credits, then a Pro option displayed at $200 per monthConfirm the selected monthly credit tier and expected debit for the actual task.
IP Patent Searching Pro pageMonthly view shows $400 per month and ten flexible search creditsThis page assigns one credit to a novelty search and two to an FTO search.
Enterprise or embedded accessEnterprise is sales-led, while the main page describes pay-as-you-go open platform accessAsk which workspaces, shared credits and API rights the order covers.

Offers displayed on 21 November 2026 in Eureka’s main pricing section and its separate IP pricing page.

Following the Pricing link in the main Eureka experience leads to its own offer. The separate IP page describes another scale of task credits and also lists a distinct patent drafting package. Both were readable when consulted. The pages do not explain how a credit on one schedule converts into a credit on the other. Ask Patsnap to identify the actual checkout route, product, renewal term and entitlement on a written order before estimating the cost of a month of patent work.

The practical unit for the latch team is a reviewed search packet. A cheap query that generates forty marginal references can consume more staff time than a costly task that retrieves three strong ones. During the trial, record the credits used by each follow-up and the minutes spent opening references. Keep experiments separate from routine searches. An open-ended investigation into terminology is valuable work, but multiplying its heaviest day across every engineer would give a misleading purchasing forecast.

05 / DistinctionsPut the comparison where the questions diverge

Clarivate’s Blueprint provides a useful comparison for a group that already licenses research and intellectual property information. Start by asking which records, identifiers and family relationships the team needs to preserve. If those are already available, compare the extra research work each option enables on the same latch, including how readily colleagues can reproduce the path from an observation to its source.

Elicit’s Blueprint is a different comparison when the immediate uncertainty is in the scientific literature. An engineer researching which test method has appeared in papers may want a structured literature review. An engineer researching who described a particular locking arrangement needs to inspect patent documents and drawings. A project can contain both questions. Giving each a separate evidence table helps reviewers see why a paper that explains an experiment and a publication illustrating a device earned their places.

06 / QuestionsWhat falls out of a neat comparison table?

A poor result can begin with a perfectly sensible translation. A component described by function in one language may be described by shape in another. Have a specialist introduce alternative terms from a genuinely relevant document and rerun a narrowed search. Record which families enter and leave the shortlist. A repeatable prompt helps identify a change in the search strategy. It cannot tell the team how much of the relevant universe was missed.

Also decide which version of an invention can leave the company. A public competitor drawing, an unpublished design revision and a supplier’s confidential attachment present different questions about upload rights. Have the information owner identify the appropriate account and permitted material before attaching files. If only a public analogue can be used in the evaluation, label it as such when interpreting the result. A good outcome on a tidy public drawing may say little about how the system handles a dense internal specification.

07 / DecisionWhich investigation should Patsnap handle first?

01

A small engineering group exploring alternatives

Try the hosted route on one public mechanism. Save the unfamiliar terms that lead to genuinely relevant documents and note which ones lead nowhere.

Take the strongest three references into the design review.
02

A patent practice comparing search tools

Identify the IP package that will actually be purchased, then check the permitted task, source quality and review effort against a completed matter.

Obtain a written credit basis before building a cost estimate.
03

An enterprise connecting data to its own portal

Specify identifiers, update cadence, access controls and the evidence engineers must be able to reopen months later. Request terms for embedded access.

Evaluate data delivery and the researcher interface separately.
What should we explore next?

A business worth understanding.

Suggest your business or one you find interesting. Tell us what you want to understand about its product, positioning, design or workflows.

Suggestions are free. Selection and publication stay with the desk.

Sources
Filed under Search & researchCompany PatsnapNot affiliated with PatsnapRequest a correctionRequest a refresh by email

Continue reading

All in this category