Automatic scanner for profitable lots for skin trading in the game CS2

Developed end-to-end automation on n8n for a CS2 skin trader: the system automatically collects new lots from CSFLOAT, enriches them with data from other markets via API, calculates key demand/liquidity metrics, and finds the most profitable opportunities. All lots and calculation results are stored in Supabase, and the user instantly receives notifications about "best finds" in Telegram and push notifications on their phone.

What has been done (functionality):

Automatic collection of new lots from CSFLOAT (via API) and data normalization.
Storage and maintenance of lot history in Supabase (a single database for search, analytics, and deduplication).
Price comparison from multiple platforms (Steam Market / CS:GO Market, etc.) for arbitrage assessment.
Calculation of metrics:
- average selling price over the last days,
- number of sales in the last 7 days,
- skin liquidity (based on sales activity/demand).
Selection rules for "profitable" lots (margin/demand/liquidity) and automatic scoring.
Notifications via Telegram API + push notifications when lots passing filters appear.
Technologies: n8n, REST API, Telegram Bot API, Supabase (DB/storage of lots).