← compare

Senior Software Engineer · Brevo · Noida, India

I make email infrastructure 10× faster.

Senior Software Engineer with 5+ years building high-throughput email delivery infrastructure — systems that ship 200M+ emails a day at 10,000+ requests per second. I lead with measurable performance: cutting tail latency, slashing database load, and rewriting hot paths for scale.

Impact, in numbers

emails / day Sustained delivery throughput across the platform.
requests / sec Peak request rate served by the send API.
p99 latency cut Tail latency reduced from 5s to 800ms via parallel processing.
fewer DB queries Daily database queries cut from 70M to 3M.

Signature work

Self-balancing worker framework

Replaced rigid, manually-tuned queues with a self-balancing worker framework that scales horizontally and migrated to it with zero downtime.

80ms → 25ms queue time 10s → 2s time-to-inbox

Batch prefetch + in-memory limits

Introduced batch prefetching and in-memory rate limits to collapse redundant database access on the hottest paths.

70M → 3M daily DB queries (25×)

Rewrote the core send API in Go

Ported the core send API from Node.js to Go, eliminating blocking I/O, lowering memory footprint, and raising sustained throughput.

Parallelized the request path

Reworked the API request path for parallel processing, cutting p99 latency and scaling the service past 10,000 requests per second.

5s → 800ms p99 latency

Incident response & reliability

Led RCA on a duplicate-delivery event affecting 12,000+ messages, fixed a 61s→2.5s latency spike in 5 days, and drove mean-time-to-detect from hours down to minutes.

61s → 2.5s latency spike

Experience

Senior Software Engineer

Brevo (formerly Sendinblue)

  • Lead a team of 4 engineers building email delivery infrastructure.
  • Designed the self-balancing worker framework now serving 200M+ emails/day.
  • Own latency and reliability for the core send path at 10k+ req/s.

Software Engineer

Brevo (formerly Sendinblue)

  • Rewrote the core send API from Node.js to Go for higher throughput.
  • Cut daily database queries 25× via batch prefetch and in-memory limits.
  • Reduced p99 API latency from 5s to 800ms through parallel processing.

Capabilities

Languages

Go · JavaScript · TypeScript

Backend & messaging

Node.js · Kafka · RabbitMQ · gRPC · REST · Pub/Sub · Distributed queues

Data stores

MongoDB · Redis · MySQL

Infra & observability

GCP · Kubernetes · Docker · CI/CD · Datadog · Grafana

Selected projects

Education

Gautam Buddha University

B.Tech + M.Tech (Integrated), Computer Science · CGPA 8.02 / 10 · 2015 — 2020