The one-page
version

Kunal Jha — Backend & Distributed Systems Engineer

Delhi, India · +91 96432 21081 · kjha38017@gmail.com · linkedin.com/in/kunaljhaworkspace · kunaljha.com

Summary

Backend engineer with 5 years of Java/Spring Boot microservices on AWS at Ultimate Kronos Group, Amazon and Accenture. Owns multi-region, event-driven systems end to end — API contracts to on-call — and wires LLMs into real infrastructure via MCP servers, agents and Claude Code skills. Latency down 25–40%, manual verification down 30–40%, reliability up ~30%.

Experience

Ultimate Kronos Group — Software Engineer 3 · Jun 2026 – Present · Delhi

• Cut legacy API latency by up to 40% — profiled hot paths (N+1 queries, redundant serialisation, chatty DB access) and introduced Redis caching for read-heavy lookups.
• Architected a support API that consolidates scattered RabbitMQ topic listeners into one observed service — topic-based routing, de-duplication and shared error handling; new topics onboard by configuration, not new listener code.
• Built MCP servers for Kubernetes, Jira and Salesforce — typed, permission-scoped tools (read-only defaults, allow-listed fields, explicit approval for writes) that let Claude query pods and logs, manage tickets and pull CRM context.
• Authored Claude Code skills: a GitHub workflow skill and a hot-patch skill that builds with Gradle and swaps fresh JARs into running Kubernetes pods and HA machines with backup and rollback — 15 manual steps became one command.
• Put Claude into the support-ticket workflow — triage, severity suggestion, runbook-grounded root-cause hints, an engineer approving every reply — and shipped an issue-context agent that turns Confluence and Jira searches into a cited brief.

Amazon — Software Engineer · Supply Chain Tech · Jul 2024 – Jun 2026 · India

• Owned multi-region, fault-tolerant microservices for logistics across the India, Brazil, UAE and KSA marketplaces — region-local processing with replicated state (DynamoDB), so a regional incident degrades gracefully; ~30% higher reliability.
• Designed the async messaging backbone on SQS / SNS — idempotent consumers, retry with back-off and dead-letter queues — so duplicate and failed events are handled by design rather than by the on-call engineer.
• Engineered the extensible APIs and data models behind the Electronic Proof of Delivery platform, cutting manual verification 30–40% for delivery-completion workflows.
• Defined versioned, backward-compatible API contracts and event schemas (Kafka) so downstream marketplace teams could integrate without coordinated releases.
• Led the migration of legacy compliance workflows into domain-split, independently deployable services — 25–35% lower latency; wrote the design docs, hardened auth boundaries (IAM) and ran a side-by-side cut-over with zero downtime.
• Improved notification reliability and observability 20–30% — structured logging, CloudWatch dashboards and alarms, retry architecture — moving alerts from "customer reported" to "caught before impact".
• Carried the on-call rotation and drove root-cause analyses for recurring incidents; prototyped AI-driven log analysis and anomaly detection to shorten debugging and resolution time.
• Brought AI-assisted development tooling into the team's backend workflow, shortening iteration time on feature delivery.

Accenture — Software Engineer · Banking · Sep 2021 – Jul 2024 · India

• Built Java/Spring Boot REST services for banking customer-onboarding journeys — KYC steps, document capture and account creation — serving high daily volumes.
• Externalised eligibility and validation rules into a rule-driven workflow engine, improving onboarding efficiency ~15% and letting business changes ship without code deployments.
• Hardened integrations with core-banking and third-party verification systems — error handling, retries and idempotency — reducing manual errors ~30%.
• Owned the automated integration test suite for critical onboarding flows and wired it into CI; recognised with Accenture's Client Value Creation award.

Technical skills

• Languages: Java
• Backend & messaging: Spring Boot, REST API design, microservices, event-driven systems, RabbitMQ, Apache Kafka, SQS / SNS
• Data: PostgreSQL, Amazon DynamoDB, Redis, Elasticsearch, MySQL
• Cloud & DevOps: AWS (Lambda, API Gateway, S3, EC2, ECS / Fargate, CloudWatch, IAM), Docker, Kubernetes, GitHub Actions
• AI engineering: MCP servers, Anthropic API, Claude Code skills, agents, RAG (pgvector), tool design & permission scoping

Education & certifications

• B.Tech, Computer Science — Guru Gobind Singh Indraprastha University, Delhi · 2021
• Anthropic — Claude with the Anthropic API · Model Context Protocol · Claude Code in Action · 2026
• Udemy — System Design (LLD + HLD) · 2024