Vishal Kumar
Experience
Projects
Skills
Recommendations
Contact
Resume
VK

Vishal Kumar

Software Engineer Platform & Infrastructure

I build platform infrastructure and distributed systems. At Atlan, I migrated 600+ production tenants to CloudNativePG with zero data loss, built Blackbird (a declarative config system managing 800-1000 tenant clusters), and productionized the Temporal workflow engine. I use Claude Code heavily as a daily driver, which compresses 2-3 months of engineering work into 2 weeks, routinely.

Active in Google Developer Groups (Bangalore, Kolkata, Durgapur) and the K8s Bangalore community.

Resume

Experience

AtlanSoftware Engineer | Platform
Nov 2024 – Present(1y 9m)
OpeninAppBackend Engineer
Feb 2024 – Nov 2024(10m)
RelectureFounding Software Engineer
Apr 2021 – Apr 2023(2y 1m)

Projects

ItelcotechFreelance
Jul 2023 – Dec 2023(2y 7m ago)

Production e-learning platform for telecom technology education. Course commerce, DRM-protected video learning, quizzes, certificates, and B2B/academic multi-tenant org memberships. Built on Next.js 14, tRPC, Prisma/Neon, and Sanity CMS, with dual-currency Stripe payments. Also designed an on-demand cloud-labs system that provisions real AWS EC2 instances per learner for hands-on 5G core network (Free5GC) exercises, with automated provisioning, idle cleanup, and minute-billed credit tracking.

Vacation ResponderBackend
Nov 2023 – Jan 2024(2y 6m ago)

Node.js CLI app that monitors incoming emails via Gmail API and automatically sends contextual responses to first-time email threads, keeping inbox organized during absence.

TaskflowFull-Stack
Jan 2023 – Mar 2023(3y 4m ago)

Collaborative project management app with multiple workspaces, drag-drop boards, team roles, and customizable boards.

InformFull-Stack
Jun 2023 – Aug 2023(2y 11m ago)

Web-based form builder allowing creation, sharing, tracking and editing forms with multiple validation types (MCQ, text, email, URL).

MakeItSafeBot
Oct 2020(5y 9m ago)

AI moderation Discord bot detecting inappropriate text and images using Perspective API and NSFW.js to keep communities safe.

All projects

Skills

Backend SystemsGo, TypeScript/Node.js, Python, REST APIs, Microservices, Express, Fastify, tRPC, gRPC
Distributed SystemsTemporal, Kafka, RabbitMQ, Kinesis, Dapr, Asynchronous Processing, Multi-tenant Architecture, Idempotent Operations
Infrastructure & CloudKubernetes, Docker, Helm, ArgoCD, CI/CD, GitHub Actions, AWS (Lambda, S3, Kinesis, Athena, Glue), Linux
AI EngineeringOpenTelemetry, Grafana, Distributed Tracing, Metrics, Alerting, Production Incident Management, Claude Code, AI Agent Development, AI Workflow Automation
DatabasesPostgreSQL, CloudNativePG, Redis, MySQL, ClickHouse, S3/MinIO

Recommendations

OpeninAppOct 2024
“I highly recommend Vishal Kumar for his exceptional intelligence and expertise in cloud technologies. He is a true problem solver with a solid understanding of scaling systems. If you're looking for someone capable of tackling complex technical challenges, Vishal is the right choice.”
AC
Amit Choudhary

DevOps & Platform Engineer

Worked on the same team
OpeninAppJul 2024
“Vishal is a detail oriented, hard working individual with a knack for simplifying complex codebases and automating repetitive tasks. I have worked on a project with Vishal and it was a pleasure working with him. He has in depth knowledge of his skill and his codebases. He would be a very valuable asset to any organisation.”
SS
Shobhan Srivastava

Software Engineer, Data Platform

Worked on the same team
ItelcotechOct 2023
“It is quite rare when we come across a great engineer like Vishal, who not only understands the requirement of a client but also delivers beyond expectations. He has great combination of technical expertise, team collaboration and business acumen. His contentiousness is second to none. His contribution to enable a learning platform in telecom industry is remarkable. Thanks Vishal !!”
RK
Rahul Kaundal

Head - Radio Access & Transport Network

Freelance Client
RelectureSep 2023
“Vishal is a go-getter. At Relecture, he built the entire tech infrastructure from the ground up. He possesses great problem-solving skills along with the right attitude to work in ambiguous situations. Vishal will be a great addition to any organisation.”
UP
Ujjwal Pushp

Entrepreneur in Residence at Antler | 2x Founder

Managed Vishal directly

Get in touch

vishalx360@gmail.com
Resume

2026 Vishal Kumar

Atlan

Software Engineer | Platform·Nov 2024 – Present(1y 9m)
atlan.com·Bengaluru, India

System Context

Series C data collaboration platform (data catalog) serving enterprise customers globally. Platform Services team owns internal infrastructure, developer tooling, and production reliability for every engineering team.

Scale

600+ production tenants. 800-1000 customer tenant clusters. Multi-service Kubernetes deployment.

Engineering Challenges

  • –Configs drifted silently across 800-1000 tenant clusters. Engineers made temporary on-call changes that never got reverted, and nobody could tell what a tenant's actual state was versus what it should have been. Incidents happened because people assumed configs were in a known state when they weren't.
  • –Legacy PostgreSQL deployments were blocking scale.
  • –Legacy Argo workflows couldn't hold the reliability bar for critical operations.

What I Built

  • –Designed and built Blackbird from scratch, a declarative config state management system in Go. Temporal-powered drift detection runs on a schedule, and an idempotent remediation engine fixes drift across 800-1000 customer tenant clusters. Engineers onboard a new service by adding one remediation class with four methods (preflight, get, main, postprocess), and Blackbird auto-registers it and starts syncing. 20-30 services now run on it as their config source of truth, and config-drift P0 outages are gone.
  • –Migrated 600+ production tenants from legacy PostgreSQL to CloudNativePG with zero data loss and zero downtime. Automated verification, retries, and observability dashboards caught problems before they reached customers, and every migration-related P0 outage disappeared.
  • –Productionized the Temporal workflow engine with the Infra team. It now automates database backups, state remediation, and background services across teams, replacing the legacy Argo processes that couldn't stay reliable.
  • –Owned production on-call across core services (Heracles Go monolith, Ladon policy engine, Redis, PostgreSQL, Kong gateway) every quarter.
  • –Built a Claude Code plugin for early incident RCA, which sped up root cause identification during on-call.
  • –Led the Query Intelligence cost reduction work on App Platform. Used Claude Code to research and ship fundamental changes to parquet parsing logic, and migrated legacy Argo workflows to Temporal-based orchestration. Data processing costs dropped.
  • –Adopted Claude Code heavily as a daily driver. Work that used to take 2-3 months now finishes in 2 weeks, routinely.

Technical Decisions

  • –Built Blackbird as a declarative source of truth (the atlan-config repo) rather than another imperative tool.
  • –Chose CloudNativePG over a managed DB for operational control.
  • –Picked Temporal over Argo for durable workflow reliability.
  • –Centralized config orchestration to kill tenant-level drift.

Impact

  • –Eliminated config-drift P0 outages across the fleet.
  • –600+ tenants migrated, zero data loss, zero downtime.
  • –20-30 services running on Blackbird as their config source of truth.
  • –Query Intelligence data processing costs reduced.
  • –2-3 months of engineering work compressed to 2 weeks via AI-native development.

Technologies

GoPostgreSQLCloudNativePGTemporalRedisKongDockerKubernetesHelmGitHub ActionsOpenTelemetryGrafanaClaude Code

Atlan

Software Engineer | Platform·Nov 2024 – Present(1y 9m)
atlan.com·Bengaluru, India

System Context

Series C data collaboration platform (data catalog) serving enterprise customers globally. Platform Services team owns internal infrastructure, developer tooling, and production reliability for every engineering team.

Scale

600+ production tenants. 800-1000 customer tenant clusters. Multi-service Kubernetes deployment.

Engineering Challenges

  • –Configs drifted silently across 800-1000 tenant clusters. Engineers made temporary on-call changes that never got reverted, and nobody could tell what a tenant's actual state was versus what it should have been. Incidents happened because people assumed configs were in a known state when they weren't.
  • –Legacy PostgreSQL deployments were blocking scale.
  • –Legacy Argo workflows couldn't hold the reliability bar for critical operations.

What I Built

  • –Designed and built Blackbird from scratch, a declarative config state management system in Go. Temporal-powered drift detection runs on a schedule, and an idempotent remediation engine fixes drift across 800-1000 customer tenant clusters. Engineers onboard a new service by adding one remediation class with four methods (preflight, get, main, postprocess), and Blackbird auto-registers it and starts syncing. 20-30 services now run on it as their config source of truth, and config-drift P0 outages are gone.
  • –Migrated 600+ production tenants from legacy PostgreSQL to CloudNativePG with zero data loss and zero downtime. Automated verification, retries, and observability dashboards caught problems before they reached customers, and every migration-related P0 outage disappeared.
  • –Productionized the Temporal workflow engine with the Infra team. It now automates database backups, state remediation, and background services across teams, replacing the legacy Argo processes that couldn't stay reliable.
  • –Owned production on-call across core services (Heracles Go monolith, Ladon policy engine, Redis, PostgreSQL, Kong gateway) every quarter.
  • –Built a Claude Code plugin for early incident RCA, which sped up root cause identification during on-call.
  • –Led the Query Intelligence cost reduction work on App Platform. Used Claude Code to research and ship fundamental changes to parquet parsing logic, and migrated legacy Argo workflows to Temporal-based orchestration. Data processing costs dropped.
  • –Adopted Claude Code heavily as a daily driver. Work that used to take 2-3 months now finishes in 2 weeks, routinely.

Technical Decisions

  • –Built Blackbird as a declarative source of truth (the atlan-config repo) rather than another imperative tool.
  • –Chose CloudNativePG over a managed DB for operational control.
  • –Picked Temporal over Argo for durable workflow reliability.
  • –Centralized config orchestration to kill tenant-level drift.

Impact

  • –Eliminated config-drift P0 outages across the fleet.
  • –600+ tenants migrated, zero data loss, zero downtime.
  • –20-30 services running on Blackbird as their config source of truth.
  • –Query Intelligence data processing costs reduced.
  • –2-3 months of engineering work compressed to 2 weeks via AI-native development.

Technologies

GoPostgreSQLCloudNativePGTemporalRedisKongDockerKubernetesHelmGitHub ActionsOpenTelemetryGrafanaClaude Code