Vikash Kumar
View resume

Software Engineer II · Backend & Platform · India

Systems that
cannot afford
to fail.

I build high-throughput backend systems and developer platforms that stay fast, reliable, and operable under real-world load. Currently at Temple; previously Zomato, Paytm, and BYJU'S.

View resume Discuss an opportunity
4Mrequests per minuteScaled Zomato's events homepage through local caching and request-path simplification.
95%fewer API errorsImproved ReportCard reliability through caching, lazy loading, and request-path optimization.
p99 200mscheckout latencyReduced a conversion-critical checkout path from one second.

Selected case studies

Evidence over adjectives.

A few examples of how I approach constrained, high-volume systems: understand the failure mode, simplify the path, and make the outcome measurable.

01

High-demand events at Zomato

ContextA homepage endpoint had to sustain dramatic traffic spikes without creating excessive infrastructure cost.

DecisionReworked the request path around local caching, payload optimization, and simpler service boundaries.

200K to 4M RPM
GoCachingPerformance
02

A checkout path built for speed

ContextThe District Events checkout workflow required a materially faster critical path as traffic grew.

DecisionOptimized backend processing across checkout, inventory, seat booking, and merchant payout workflows.

p99: 1s to 200ms
BackendAPIsMongoDB
03

A platform for predictable releases

ContextFirmware rollouts needed durable workflows, clear rollout controls, and safe rollback decisions.

DecisionDesigned a release-management platform with Temporal orchestration and policy-driven promotion rules.

Reliable progressive delivery
GoTemporalPostgreSQL

Experience

Built across products where reliability is the product.

01

Temple

Software Engineer II

Release management and observability platforms that let product teams deploy and operate with confidence.

GoPostgreSQLTemporalAWS
2026 - now
02

Zomato

Software Engineer II

Core systems for brand onboarding, taxation, checkout, inventory, and high-demand event operations.

GoMongoDBKafkaRedis
2024 - 2026
03

Paytm

Software Engineer

IPL-scale ticketing traffic, payments performance, and database migration workflows.

BackendQueuesMongoDB
2023 - 2024
04

BYJU'S

Software Engineer

Event-driven communications and data-platform integrations for students and teachers.

AWS SQSSalesforceAPIs
2022 - 2023

Working toolkit

From the critical path to the operating model.

Systems
Distributed systems, event-driven architecture, API design, release orchestration
Platform
AWS, Docker, Linux, Temporal, Kafka, Flink, GitOps
Data
PostgreSQL, MongoDB, DynamoDB, ClickHouse, Redis
Observability
Datadog, Grafana, Prometheus, OpenTelemetry

Notes on engineering

Writing about the decisions behind the systems.

Architecture, distributed systems, and the principles that make complex systems easier to reason about.

01Why Kafka had to existFirst principles 02The architecture of a distributed logSystem design 03Kafka in productionImplementation

Open to SDE II opportunities

Building something that needs to hold up?

I'm interested in backend, infrastructure, and platform roles where scale, reliability, and engineering judgment matter.

vikash.kumar.186057@gmail.com