DAT Streams Millions of Staff Messages Through Neon

Neon published a case study on how DAT's database layer supports mission-critical field operations for more than 16,000 Lahore Waste Management Company staff. Every incoming message from WhatsApp, Telegram, or WeChat triggers several background jobs and audit-log writes, adding up to roughly 10 million database transactions a month.

“Our system can't afford any downtime,” said Ahsan Nabi Dar, CTO and co-founder of DAT, describing why message reliability is treated as mission-critical rather than best-effort.

DAT splits reads and writes across Neon's primary instance and read replicas, letting ingestion continue uninterrupted while supervisors get real-time reporting through dashboards built on Retool. The database also scales automatically during demand spikes, such as multi-day cleanup drives, and scales to zero during quiet periods to keep costs down.

DAT also uses Neon's branching feature to test schema and feature changes against production-like copies of the database without duplicating infrastructure or committing to fixed instances ahead of time.