At Motorq, I build and operate backend features for a connected-vehicle data platform processing 250B+ vehicle signals each month. I work in a cross-functional pod that owns ingestion across India and the US.
I’ve delivered privacy-policy APIs for Motorq’s Canada expansion, compliance synchronization between recovery-vendor and customer environments, and a Stellantis video-event stream. I take work from requirements and design through observability, deployment, and production ownership.
I improved reliability by tracing enrollment-processor memory spikes from 2 GB to incorrect symbol filtering, reducing post-fix memory to roughly 150 MB. I also removed duplicate ingestion filtering, saving about $72K annually across six environments, and built an AI bug-fixing agent whose PRs achieved a roughly 60% merge rate over six months.
I led an evaluation of a Kinesis-to-Pulsar ingestion migration, found correctness issues, and contributed a corrected fix merged upstream into Apache Pulsar. I build for concurrency, fault tolerance, and performance, validating prototypes above 1M records per minute with OpenTelemetry and Grafana instrumentation.
