Pull
Retrieves the registration's maintenance filings and specimen images from the USPTO's TSDR system, document by document, within the API's rate limits, and caches them.
Private beta · Investor preview
Find the weak registrations before the opposition does.
MarkBuster is software that audits the maintenance record behind a trademark registration: what was filed to keep it alive, what it actually showed, and whether it holds up. It turns weeks of manual docket reading into a ranked, documented list of leads for an attorney to review.
- In beta nowrunning on real portfolios
- Under NDAcase studies for investors
- Public record onlyUSPTO TSDR and TTAB data
Beta results
First beta run: one prominent law firm's portfolio
We pointed the engine at the trademark filings of a large, prestigious U.S. law firm. We are not naming the firm at this stage.
3,946registrations indexed by attorney of record
853maintenance filings pulled and read
1,174TTAB proceedings mapped to those registrations
59registrations flagged for review
Figures from the beta run as of November 9, 2026. "Flagged" means a screening lead that a person reviewed against the source filing; it is not a finding by any tribunal.
Why it matters
The flagged registrations belong to well-known brands, and the filings that kept them alive were signed by the firm's own attorneys. Those same registrations have been used to oppose small companies. In our assessment, the findings could carry potentially serious exposure for the firm and for the high-profile clients it signed for. The result is opinion, drawn from the public record, and every lead links to the USPTO source so a reader can check it.
We are building case studies from this run now. They will go into investor packages that are released only under a non-disclosure agreement, in stages.
The problem
Registrations get used as weapons. Almost nobody audits them.
A registered mark has to be kept alive with filings that say the mark is still in use, backed by a specimen. Those filings are public, but reading them one at a time across a portfolio is slow, expensive and easy to skip.
The result: when a company opposes your trademark, it leans on a stack of registrations that no one on your side has checked. A weak one can change the case. Finding it by hand costs a litigator days, and by then the budget is spent.
What a manual review looks like today
- Pull each registration's file history one document at a time
- Open each specimen image and judge it by eye
- Remember what was filed in earlier cycles
- Write it up in a form someone else can verify
MarkBuster does the first three automatically and does the fourth the same way every time.
The product
An evidence engine, not another search box
Trademark search tools tell you what is registered. MarkBuster examines the maintenance evidence behind it.
Read
Runs text recognition on every specimen to catch markers such as "only 1 left in stock", "buy used" and a shopper's delivery address, and checks whether the page shows the registrant's own use.
Compare
Crop-aware image comparison across filing cycles finds a photograph that was cropped, enlarged and filed again, a case ordinary image matching misses.
Map
Links each registration to the TTAB proceedings in which it was pleaded, so you see where a weak registration has actually been used.
Rank
Scores each lead on how directly the filing contradicts the claim of use, with the basis and the caveat written next to every score.
Prove
Exports an evidence package: source images, SHA-256 hashes, retrieval times, comparison settings and reviewer decisions, so the work can be reproduced.
Example leadidentifiers removed
- Filing
- Sections 8 & 9 renewal, signed by counsel
- Specimen filed
- One photograph of retail packaging, no longer sold
- Earlier cycle
- Same photograph, filed ten years before
- Crop-aware match
- 0.997 correlation at one uniform scale (unrelated images score 0.5 to 0.9)
- Status
- Flagged for attorney review
Illustrative. Drawn from the beta run with identifying details removed. Real output links each line to the USPTO document.
Reverse enforcement
Enter a TTAB opposition or cancellation number. MarkBuster pulls the registrations the other side pleaded, retrieves their maintenance history and returns a ranked report on each one, before you draft a defense.
Portfolio self-audit
A corporate IP department screens its own portfolio for questionable past filings and fixes them before an adversary finds them.
Output is a set of leads for an attorney to review. A person reviews each one before anything leaves the system.
Market and model
Sold first to the people who litigate
Trademark litigation firms are the first customer. Corporate IP departments doing internal risk audits are the second. We are not competing with search and docket-monitoring services; we sell the investigation they do not offer.
| Offer (hypotheses to test) | Starting price |
|---|---|
| Single registration automated screen | $99 to $199 |
| Litigation research package, 10 to 25 registrations | $1,500 to $4,000 |
| Solo or small-firm subscription | $399 / month |
| Litigation team subscription | $1,250 / month |
| Enterprise portfolio screening | from $3,000 / month |
Pricing is our working hypothesis, not validated willingness to pay. Our first step is five supervised research packages with real attorneys, to learn what they value before we build billing. Usage credits, not unlimited scans.
Roadmap
From working engine to a product firms can trust
- Database as source of truth. Permanent finding IDs, every note and correction preserved, spreadsheets generated as exports.
- Provenance on every result. Source hashes, retrieval times, comparison settings, reviewer decisions and algorithm version.
- Case intake. Enter a proceeding number; get the pleaded registrations and a ranked report.
- Read-only web interface. Search, sortable leads, side-by-side image viewer, links to the USPTO record.
- Attorney pilot. Three to five IP attorneys on live matters; we measure precision, usefulness and what they will pay for.
Our own estimate is a few focused weeks to an internal proof-of-concept interface; a secure multi-client service is a larger build because of data isolation, evidence provenance and quality assurance. Targets, not commitments.
The raise
Exploring investors, and possibly crowdfunding
We are talking with potential investors and are considering a crowdfunding round alongside them. Size, structure and terms are not set; we will settle them with securities counsel first.
Investor materials, including case studies from the beta, will be released in stages and under a non-disclosure agreement.
This website is not an offer to sell, or a solicitation of an offer to buy, any security. Any offering will be made only through the appropriate exemption or registration and its own documents. Nothing here is investment advice.
What the funding is for
- Moving the engine onto a database with full provenance
- A secure, multi-client web product
- Attorney pilots and quality review of the output
- Legal, security and data-access compliance
Plain risks
What an investor should weigh
Early stage
MarkBuster is in beta. One portfolio run is not proof the product works across matters, and nobody has yet paid for it.
Leads, not verdicts
Screening misses things and over-flags others. Output needs an attorney's review, and its precision is still being measured.
Data access
The product depends on USPTO data and its published rate limits and terms, which can change.
Disputes
Findings about a named firm or its clients can draw legal challenges. That is why we are not naming the firm and why we rely on public-record sources.
Founder disclosure
The founder's company is a party to a pending trademark proceeding against a large company, and the first beta run grew out of research done in connection with it. Details are available under NDA.
Competition
We have not found a product that combines crop-aware specimen comparison, maintenance screening and TTAB mapping. A search is not proof that none exists.
Request the investor brief
Tell us who you are and what you would like to see. We will reply with the brief and, where appropriate, an NDA for the case studies.