Real-Time Smart-Money Alerts for a Crypto Trader

The problem

A US-based trader tracks a set of smart-money wallets. The signal they wanted: the moment any of those wallets shows up as the funding source behind a holder of a newly trending Solana token on GMGN.ai — something that requires watching two pages at once and cross-referencing them continuously. No human can do that all day.

The build

I built a two-browser architecture. The first browser watches GMGN’s “Almost Bounded” and “Migrated” token tables, detecting new and re-ranked tokens every two seconds. Each change is queued for a second, background browser that opens the token’s holders page and extracts every holder’s funding address.

GMGN’s holders table is virtualised — only the visible rows exist in the DOM — so the scraper auto-scrolls, deduplicates as rows render, and recovers from the stale-element errors that re-rendering causes. Onboarding pop-ups are dismissed automatically, and the monitoring page refreshes itself every five minutes to stay current.

Matches against the trader’s watchlist trigger an audio alert and are logged with timestamps to CSV. Delivered as a standalone Windows executable — no Python install needed.

The result

Round-the-clock smart-money surveillance with instant alerts, delivered in six days. This was the client’s third project with me.

“My 3rd project with Muneeb and as always he is very smart and fast in getting the job ✅” — via Fiverr