LightfieldAI-native CRM that builds itself and does work for you
Promoted
Real-Time Analytics Pipeline
Real-time analytics: FastAPI, Kafka, TimescaleDB
11 followers
Real-time analytics: FastAPI, Kafka, TimescaleDB
11 followers
Enterprise open-source real-time analytics pipeline: 10K+ events/sec, sub-50ms latency. FastAPI async API, TimescaleDB time-series, Kafka streaming, Redis caching, WebSocket live, Prometheus/Grafana monitoring, OpenTelemetry tracing. Anomaly detection, K8s auto-scaling, 150+ tests. Ideal for IoT, fintech, observability.
Set it up over the weekend and was impressed by how smoothly the WebSocket feed handled 8K events/sec without dropping a frame. The Prometheus dashboards are genuinely useful out of the box, which saved me a ton of wiring time.
Spent an afternoon poking around the repo and the WebSocket live feed felt genuinely snappy, even when I hammered it with a few thousand synthetic events. Love that TimescaleDB + Kafka combo is wired up cleanly out of the box.
Spent an afternoon poking around the repo and the FastAPI setup is genuinely clean, async endpoints just work without the usual boilerplate. The Prometheus metrics out of the box saved me an hour of wiring.
Spun up the demo stack over the weekend and the sub-50ms latency claim actually holds up under load, which I was skeptical about. The WebSocket dashboard updates felt snappy even when I pushed synthetic traffic near the 10K/sec mark.
Deployed a sample IoT workload against it and the end-to-end latency held steady around 40ms even when I pushed past 8K events/sec, which honestly beat what I expected from a fully open-source stack.