Public Vector is a consumer-protection intelligence service for plaintiff's counsel. Deploying autonomous agents across the consumer economy allows us to track marketing claims, pixel and SDK deployments, pricing flows, consent UX and more at scale. We cross-reference our findings against scientific literature, legal expertise, and the docket to surface viable cases, and package collected evidence for counsel to review — in the dashboard, or as sealed, exclusive case packets on the marketplace.
Vigilance never sleeps: three monitoring layers sweep the consumer economy on a continuous cycle.
Autonomous browser agents continuously sweep consumer-facing sites — reading marketing claims at scale, exercising consent banners, walking pricing flows, and recording the trackers that fire — so violations are observed as they happen on the defendant's own pages, not reconstructed after the fact.
We continuously monitor where consumers already report harm — regulator complaint databases, consumer-protection boards, forums, and review sites. A violation only advances when independent complaint volume corroborates it, so evidence arrives with putative class members already on the record.
Literature monitoring flags products whose claims new research contradicts. Daily grievance time-series spot accelerating harm categories before they make news. And every candidate clears a prior-litigation screen — federal docket plus the open web — so what surfaces is first-mover.
Every defendant in the dashboard carries an arbitrage gap score between zero and one — the distance between what the site does and what its policies actually disclose. Filter, sort, drill in, and pull a sealed PDF on any defendant you want to pursue.
Each marketplace listing is a defendant our agent has already verified live: verbatim representations captured, network traffic recorded, wrap classification done, sealed PDF assembled. The first buyer claims the listing exclusively — once sold, it's gone.
Surveillance spots the violation. This pipeline turns it into evidence.
Our agent drives the defendant's site like a real user — submits forms, hits checkout, exercises consent banners — while a network listener records every tracker that fires and a representation extractor pulls verbatim marketing language.
We cross-reference the captured tracking behavior against the defendant's own privacy policy, ToS, and cookie disclosures. The result is an arbitrage score: the documented distance between what the site does and what it tells consumers.
Every artifact — screenshots, HAR captures, policy snapshots, case-law alignment — is hashed and bundled into a sealed PDF with FRE 901 chain of custody. Drop it straight into a complaint or class-cert backup record.
Where our data comes from. Everything in a report is first-party: our own agents drive the defendant's live public site and capture the network requests, marketing representations, and consent UX exactly as they served — nothing licensed, nothing second-hand, every artifact hashed at capture. Scope details in the FAQ.
New here? The full walkthrough covers Quick Scan, the Dashboard, the Web Agent, and a sample sealed PDF.
— The Public Vector team