Where to Store the Calls: Schema Design and Deduplication
🧭 Designed the call-log table in PostgreSQL. The big concern was duplication: counting spam and auto-forwards as one call or many changes everything. We split call records from coin metadata.
To analyze the calls we gathered, we first had to store them well. We chose PostgreSQL as the database and designed a call-log table that records, for each call, which coin was named, in which room, and when. The central concern was duplication: when the same person posts the same coin repeatedly, or a bot auto-forwards it, treating that as one call or many changes the statistics entirely. We also kept a separate table for each coin's market information, splitting the call records from coin metadata. This structure became the starting point for every tally and analysis that followed.
Today's Top 5 Calls
- 1STACCANAmemeATH $454.4K35 calls· Widest reach with 35 calls across 14 rooms, ATH market cap $454.4K.
- 2SUEAAUANmemeATH $425.1K33 calls· 33 calls across 12 rooms, ATH market cap $425.1K.
- 3DMCmemeATH $251.4K22 calls· 22 calls across 11 rooms, ATH market cap $251.4K.
- 4SPIRIT2.0cultureATH $364.4K20 calls· 20 calls across 6 rooms, ATH market cap $364.4K.
- 5ECASHfinanceATH $624.9K19 calls· Fifth on calls (19) yet the day's largest by size at a $624.9K ATH market cap.