Case study / 01
SeatSecure
One event page, every marketplace. SeatSecure turns scattered resale listings into a single live price view for the secondary ticket market.
The problem
Resale prices for the same seat are spread across StubHub, SeatGeek, Vivid Seats, Gametime, and Viagogo. Comparing them means five tabs, five fee structures, and five different views of the same event — and the numbers move while you look.
SeatSecure collapses that into one canonical event page with a live price range, updated from the source sites.
What we built
Marketplace fan-out
A FastAPI service that takes an event URL and scrapes five resale marketplaces in parallel, streaming listings back as they arrive.
Antibot transport
Browser-grade TLS and header ordering, in-process AWS WAF solving, DataDome handling, warm session pools, and proxy rotation.
Price lake
Every run writes normalized snapshots to S3 — event, venue, price range, and per-marketplace breakdown — so the site can render current numbers and keep a history.
Event surface
Canonical event pages with live ranges and OG cards that re-render when the price changes, plus automated social cards for distribution.
How it works
StubHub anchors the event metadata and seat map. Four more marketplace workers run concurrently, each with its own session and proxy, and push results into a shared queue. The browser receives the merged view over SSE, so the page fills in as sources respond instead of waiting for the slowest one.
Each scrape also lands in the price lake. That separates “what is true right now” from “what the page shows”: the event surface can be regenerated from stored snapshots, and the OG image is versioned by price range so a shared card never silently goes stale.
Stack
More case studies are on the way. Want something like this built? Get in touch.