Kaustubh
Alandkar

|

I build real-time, event-driven systems that handle production workloads.

01 / about

What I build

I build backend systems that ingest, process, and deliver data reliably in production.

My work includes real-time pipelines, backend services, and event-driven systems used in industrial and high-throughput environments.

I focus on software that stays reliable under load, is easy to evolve, and solves real operational problems.

Systems Over Features

Think in terms of system behavior, data contracts, and failure modes, not just feature delivery.

Built for Load

Design for throughput, correctness, and stable behavior under real production traffic.

Operational Reliability

Prioritize fault isolation, graceful degradation, and systems that stay dependable under pressure.

Evolving Architecture

Build modular backends that are easier to scale, test, and change as requirements grow.

02 / tech stack

What I work with

Tools and technologies I rely on to build production-grade backend systems.

Languages

TypeScriptJavaScriptJavaPythonC++

Backend & APIs

Node.jsExpress.jsSpring BootFastAPIREST APIsMicroservices

Messaging & Real-Time

Apache KafkaWebSocketsSocket.IOMQTT

Databases & Caching

PostgreSQLMySQLMongoDBRedis

Cloud & DevOps

AWS EC2DockerCloudflare WorkersCloudflare D1PM2CI/CD (GitHub Actions)

Frontend

ReactNext.jsTailwind CSS

Testing & Quality

Jestk6SonarCloud
03 / projects

Featured projects

Production-style systems that showcase backend depth, distributed architecture, and real-world reliability.

View all on GitHub
04 / engineering mindset

How I think about systems

The engineering convictions that shape how I design, build, and reason about backend systems.

Fault Tolerance

Design for Failure First

If the external dependency goes down, if the message duplicates, if the connection drops — what happens? I design around these cases before thinking about the happy path.

Architecture

Clear Boundaries Enable Scale

Clear service boundaries, explicit data contracts, and isolated failure domains make systems independently scalable and safe to evolve. Monoliths are fine — until they're not.

Correctness

Idempotency Is Not Optional

In distributed systems, "at-least-once" delivery is the reality. Every write path needs to handle retries safely. I build deduplication and idempotency in from the start.

Performance

Observe Before You Optimize

I load-test before tuning, instrument before diagnosing, and measure before claiming improvement. k6 load tests and SonarCloud analysis are part of my build process.

Event-Driven

State Reconstruction Over State Storage

Event logs that can replay. Redis-backed hydration that survives restarts. The system's current state should be derivable from its event history — not assumed to persist.

Craft

Clarity Scales Better Than Context

Code is maintained longer than it is written. Clarity, explicit contracts, and test coverage are acts of engineering respect.

05 / writing

From the blog

Writing about backend systems, distributed architecture, and engineering decisions.

All posts on Dev.to

No articles found. Check back soon.

____

Let's work together

Open to discussing backend systems, distributed architecture, and practical engineering problems.

Start a conversation

For opportunities, collaborations, or thoughtful engineering discussions.

0/1000

Clear context helps me respond faster.