Building an Analytics Engine Using MongoDB and Go
summary
Builds a realtime relevancy engine that loads arbitrary offer and transaction feeds into MongoDB and runs rule-based targeting from a Go-based rules engine; uses Iron.io for scheduled feed ingestion and bgo for the web API. Rules execute as aggregation-pipeline queries (pre-2.6 limitation required template-based joins across results), update at runtime without deploys, and select offers when aggregated sums or boolean tests match. Demo shows live queries, on-the-fly rule edits, and claims handling spikes like 10,000 users or feeds without rebuilding schema.
AI-written summary · about
chapters (24)
- 0:00Introduction
- 0:12Power Wallet
- 3:43Walmart Story
- 4:14User Data
- 9:25MongoDB
- 9:59MongoDB Collections
- 10:33Aggregation Pipeline
- 11:12Dynamic Rules Engine
- 12:09MongoDB Aggregation
- 12:45Variable Rules
- 13:56Pipeline Queries
- 15:06Aggregation
- 16:58Transform
- 17:12Demo
- 18:33Target Offer
- 20:13Lipid Feed
- 20:27Pipeline Query
- 20:44Group Query
- 21:05First Match
- 22:12Offer Pipeline
- 22:51Rule Change
- 23:10Offer Targeting
- 25:41Code Generation
- 26:00Scaling
related talks
45mReal Time Analytics with MongoDB
26mBuilding generative AI–enriched applications with AWS & MongoDB Atlas
28mBuilding a Recommendation Engine for Bus Search Based on Real-Time Feedback of Users
28mBuilding AI Applications in Go
28mManaging and Scaling Real Time Data Pipelines using Go
01h 27mFlash Talks
24mMastering Data Complexity with MongoDB's Unified API
46mStreaming ETL on the Shoulders of Giants