OPEN TO OPPORTUNITIES — AVAILABLE FROM OCT 2026

Vignesh
Krishnamurthy

Principal Engineer
Backend Platforms & Distributed Systems

I design and run the systems behind streaming platforms used by millions — authentication, billing, catalogue, and the infrastructure that keeps them online. 13 years in production. Currently extending that discipline into agentic AI.

// career.timeline

13 years in production

Click any company to see what I built there.

// capabilities.list

What I build

AUTH & IDENTITY

Authentication platforms at scale

Built and ran the IDP for Hubbl's full platform — OAuth2, JWT, API-key strategies handling 5,000 req/sec at peak, 1,500 sustained. Separately, designed a token-exchange service that cut Auth0 cost by 30% and load by 50–60% through clean authn/authz separation.

EVENT-DRIVEN SYSTEMS

Kafka-driven subscription & billing

Architected the subscription management backbone for Binge — Kafka-driven, async, engineered so a user clicking "subscribe" gets confirmation within 30 seconds end to end, every time, at production scale.

OBSERVABILITY

Production visibility & incident response

Built the New Relic observability suite from scratch — dashboards, synthetic monitors, alerting wired into Slack — delivering incident notifications within 5 minutes of detection. Stood as primary escalation point for production incidents across the platform.

PLATFORM MIGRATION

Cloud-native infrastructure

Led the migration of 20+ services from EC2 to Kubernetes for Kayo Sports, giving the platform auto-scaling, faster rollbacks, and the elasticity to absorb live sports traffic spikes without manual capacity planning.

SECURITY

Security partnership, not just patching

Partnered directly with external pen-test teams, providing API specs and platform context for annual assessments. Co-ordinated remediation of every critical finding — zero carry-over between cycles, every year.

AGENTIC AI

Now building: AI-native tooling

Extending 13 years of backend discipline into agentic systems — Spring AI, LangChain4j, and Ollama, applied to real problems rather than demos. See the projects below.

// stack.deps

Stack

Languages

  • Java 17+
  • Kotlin
  • Python
  • JavaScript

Frameworks

  • Spring Boot 3.x
  • Spring Security
  • Spring AI
  • LangChain4j

Cloud & Infra

  • AWS (EC2, S3, SQS/SNS, Step Functions)
  • Kubernetes
  • Docker
  • Terraform

Data & Messaging

  • Kafka
  • PostgreSQL
  • DynamoDB
  • Redis

AI & Agentic

  • Spring AI
  • LangChain4j
  • Ollama
  • MCP

Observability

  • New Relic
  • Distributed tracing
  • SLO/SLA design
  • PagerDuty-style on-call
// projects.personal

Side builds

What I work on after hours — mostly an excuse to get hands-on with agentic AI.

Expense Analyser

ACTIVE

Spring Boot 3 · React · Ollama · LangChain4j · PostgreSQL

An AI-powered expense tool that runs a local LLM end to end — natural-language categorisation, a ReAct agent loop for multi-step reasoning over financial data, and early RAG patterns for personalised insight generation. No cloud LLM calls; everything runs on Ollama.

HomeTable

ACTIVE

Spring Boot 3 · Kubernetes · PostgreSQL · Kafka · AWS

A hyperlocal home-food marketplace concept — event-driven microservices backend, CQRS and outbox pattern across seller, order, and payment domains, deployed on Kubernetes.