CV
Viveak Vadivelkarasan
Motion
London, UK

Latest project · current role

Designing delivery for ~1M events a day.

Role
Tech Lead, External APIs
Period
Jul 2025 - Present
Attribution
Current role. Employer deliberately not named on this site.
Supported outcome
~1M events/day through the event-driven delivery service

Overview

My latest project is an external API platform: the integration surface for reservations, availability and pre-booking, used by 5 major third-party clients with more onboarding. Every external client integration depends on the event-driven webhook delivery service, which carries around one million events a day, and a high-volume data service was re-platformed from AWS Lambda onto EKS along the way.

  • AWS EventBridge
  • BullMQ
  • Redis
  • PostgreSQL
  • EKS
  • HMAC
  • Datadog

My responsibility

I lead a team of 9 engineers across multiple parallel workstreams, working closely with product, delivery and the technical architect, and I designed and built the event-driven webhook delivery service behind every external client integration.

  • Technical lead for a team of 9 engineers across parallel workstreams, owning architecture and engineering standards for the platform and working closely with product, delivery and the technical architect.
  • Designed and built the event-driven webhook delivery platform behind every external client integration.
  • Authored the platform's push-to-pull integration strategy and its four-phase rollout.
  • Own the shared authentication and authorisation tooling used across the external APIs estate.

Supported technical work

  • Webhook delivery at ~1M events/day on AWS EventBridge and BullMQ/Redis.
  • HMAC-signed payloads, per-subscription circuit breakers, rate limiting and exponential-backoff retries.
  • A Postgres-backed dead-letter queue for deliveries that exhaust their retries.
  • Re-platformed a high-volume data service from AWS Lambda to a long-running EKS service, removing timeout ceilings and cold starts while adding back-pressure handling, resumability and full observability.
  • Push-to-pull integration strategy: operation-status endpoints with Retry-After and ETag semantics, rolled out in four phases.
  • Engineering standards for the platform: outbox publishing, idempotent mutations, correlation-ID tracing and Datadog observability.
  • Introduced team-wide use of AI coding agents and agentic workflows across design, delivery and code review.

Supported outcome

  • ~1M events/day handled by the webhook delivery platform.
  • Integration cost for clients reduced through the push-to-pull strategy.
  • Timeout ceilings and cold starts removed from the re-platformed data service.
  • Throughput raised without lowering standards after introducing agentic workflows.

Conceptual diagram

Event sourcereservation.confirmedQueueordered by ageWorkerWorkerSubscriber Aclosed · 200Subscriber Bhalf-open · probeSubscriber Cclosed · 200Dead-letter queue1 parked · replayretries · backoff · circuits
Conceptual illustration. A resilient delivery pipeline: events fan out through a queue and workers to subscribers, each with its own circuit and retry policy, with exhausted deliveries parked in a dead-letter queue. It is not a diagram of any employer's systems.

What this page does not claim

The claims above are limited to what the CV supports. To keep them honest:

  • The interactive lab on this site is an independently authored educational simulation of the same delivery patterns. It uses synthetic data and is not connected to employer systems.
  • The employer is deliberately not named on this site. No customer data, internal architecture or non-public metrics are published here.