AboutExperienceProjectsOSSSkillsWritingQ&A

Software Engineer

Krish Srivastava

Software Development Engineer I @ PhysicsWallah

Building PhysicsWallah’s Internal Developer Platform and platform tooling: onboarding, workflows, load testing, and reliability at scale.

Email meResumeGitHubLinkedIn
LatestPratibimba: a reflection of your app on every pull request
Krish Srivastava
Explore

About

Hi, I'm Krish.

I'm a software engineer focused on backend systems, internal platforms, and developer experience.

At PhysicsWallah I help build and scale the Internal Developer Platform in Go, GraphQL, and PostgreSQL. It unifies infrastructure workflows across teams: service onboarding, GitLab access, Kafka provisioning, RBAC, and self-service ticketing, with better visibility for the platform team and less manual ops work.

I've designed a modular workflow engine for platform operations, migrated organizational load testing from JMeter to k6 with Jenkins automation for Vishwas Diwas, and contributed to a MongoDB to Aerospike migration POC plus API consolidation into a dedicated microservice.

I also built AI-assisted incident investigation for the Devastra platform (Zenduty, Prometheus, Elastic APM, Kubernetes, RAG over Confluence, Slack, OpenAI SDK) and internal Go tooling such as a k6 to Postman converter, with plans to open-source more of it.

Outside of work I contribute upstream in Go (redis/go-redis, hashicorp/go-set) and write on this blog about what I build.

Education

B.Tech Computer Science & Engineering

SRM University, AP · 2021 – 2025 · CGPA 7.99

✦

Experience

Where I've built

Platform engineering, reliability, and developer experience at scale.

Mar 2025 – Present

Software Development Engineer I

PhysicsWallah · On-site

  • ▸Built and scaled the Internal Developer Platform (Go, GraphQL, PostgreSQL): unified infrastructure workflows, automated onboarding, GitLab, Kafka, RBAC, and self-service/ticketing with stronger platform-team visibility
  • ▸Designed a modular self-service workflow engine so teams provision infrastructure and access through standardized, RBAC-controlled flows instead of manual platform intervention
  • ▸Migrated organizational load testing from JMeter to k6; built Jenkins-based load-test workflows and supported high-scale testing for Vishwas Diwas
  • ▸Contributed to MongoDB to Aerospike migration for a read-heavy workload (POC validation) and consolidated APIs from multiple services into one dedicated microservice
  • ▸Built AI-assisted incident investigation for Devastra: Zenduty, Prometheus, Elastic APM, Kubernetes, service docs for L1 responders, plus RAG over Confluence and Slack via the OpenAI SDK
  • ▸Developed internal Go libraries and developer tooling across microservices, including k6 script to Postman collection conversion (planned OSS release)

Projects

Selected work

Systems I've designed and shipped — from distributed storage to full-stack products.

Open Source

Contributions

Patches and improvements upstream in the Go ecosystem.

redis/go-redis

Cluster routing optimizations

Refactored cmdFirstKeyPos to use cached COMMAND INFO for cluster key routing, added a lock-free Peek() method, and cut redundant lookups across pipeline and multi-shard execution paths.

View Pull Request
hashicorp/go-set

TreeSet.EqualSliceSet

Optimized EqualSliceSet by replacing unnecessary Red-Black tree allocation with sort-and-compare, reducing heap allocations and improving equality-check performance.

View Pull Request

Skills

Tech stack

Languages, platforms, and tooling I work with day to day.

Languages

GoJavaScriptTypeScriptSQL

Backend & Infra

gRPCRESTGraphQLEntRedisPostgreSQLClickHouseKongNginx

SRE & DevOps

LinuxDockerKubernetesAWSArgoCDJenkinsGitLabELKGrafana

Other

MachineryPrometheusDistributed systemsObservabilityLoad testing

Writing

Latest from the blog

Essays on code, systems, and whatever else I'm thinking about.

View all posts
Analytics Without Google Analytics (or Any Third-Party Tracker)
Featured

tech

Analytics Without Google Analytics (or Any Third-Party Tracker)

Building First-Party Analytics: No Google, No Plausible, No Cookie Banner I run my blog on Next.js, Go, Postgres, and Redis. No third-party trackers. Here's how a fire-and-forget event pipeline (Browser -> Go API -> Redis queue -> Postgres) gives me page views, scroll depth, read time, and referrers without shipping a single tracking script to readers. Plus: the PgBouncer + prepared statements bug that took down my overview endpoint, and how I built email notifications.

August 8, 2026·3 min read

For hiring teams

Questions from recruiters, hiring managers, and future teammates

Short answers I give in screens and onsite loops. Expand any question below.

I'm Krish, an SDE at PhysicsWallah on the backend platform team. Most of my day goes into the Internal Developer Platform: Go services, GraphQL APIs, and PostgreSQL, plus the workflows around GitLab, Kafka, access control, and self-service onboarding for other engineering teams. I like work where the output is other people shipping safely and quickly. I also contribute to open source in the Go ecosystem and write here when I've learned something worth passing on.