Real-time data at scale, and keeping it honest

Software Engineer · C3 AI

Latency is a design problem, not a hardware one.

Placeholder for the full write-up. Tell the real story here: the real-time sensor-data platform processing millions of readings a day, why dashboards were drifting, and how you traced it back to the ingestion path.

The pipeline redesign

Go deep on the low-latency distributed ingestion pipeline — the 90% speed-up, the deduplication engine for mismatched ID schemes, and how zero-downtime upgrades across multi-tenant deployments actually shipped without a single escalation.

Close with what you'd tell your past self before starting.