Headline

Backend Engineer | Go, Python, Node.js | Distributed Systems | AI Infrastructure | LLM Integrations | Kafka | AWS | Microservices | PostgreSQL | Redis | Docker

About

Architected a subscription microservice supporting 10M+ integrations, built distributed workflows, and scaled APIs to 1M+ daily requests — I take backend systems from 0→1 to scale. I take high ownership across design → implementation → scale. I’ve decomposed tightly coupled systems into independent microservices, reducing failure blast radius and accelerating feature delivery, enabling faster customer onboarding and improved production stability. My work has supported systems used by 400+ enterprise customers and helped convert pilots into paid B2B revenue by improving reliability and automation success. Comfortable operating in small teams and ambiguity — shipping MVP features quickly, then evolving architecture to support growth. I focus on async workflows, queue-driven microservices, multi-tenant SaaS design, and high-throughput APIs built with observability and failure recovery from day one. 🚀 Highlights • Scaled APIs for 1M+ daily requests (800ms → 300ms latency) • Improved job success rate 40–50% via distributed coordination & recovery • Decomposed monolith into microservices to improve velocity & isolation • Reduced MTTR by 30% with observability improvements • Systems used by 400+ enterprise customers • OSS tools adopted by 170+ developers 💡 Tech: Node.js, Go, Python, PostgreSQL, Redis, Docker, AWS, BullMQ, Distributed Systems

Location

Bengaluru

AI stack

claude code

claude

codex

cursor

notebooklm

Work experience

Software Developer ( Backend ) at Josys, Jul 2025 - Present · 1 yr 3 mos, Bangalore Urban, Karnataka, India, • Built a custom SaaS integration automation platform from scratch, enabling enterprise customers to sync users from applications without public APIs into Josys Core. Introduced a new product capability that opened conversations with 15% of previously unreachable accounts . • Expanded Josys's integration catalog by 60 SaaS applications through browser automation, converting unsupported tools into fully supported user-sync workflows and closing high-priority customer integration gaps. • Designed a resilient workflow orchestration platform using BullMQ and Temporal with tenant isolation, automatic retries, crash recovery, session rehydration, and self-service remediation, reducing workflow failures by 40% and eliminating 10–12 monthly support escalations. • Built the LLM integration layer for a SaaS risk scoring platform using Python and OpenAI, enabling automated risk assessment from vendor, compliance, security, and GenAI signals. • Implemented retry handling, Pydantic-based JSON validation, and fallback logic while separating evidence collection from score calculation, reducing repeated LLM calls and improving explainable risk scoring. • Built automated SaaS governance workflows that detected unapproved applications and triggered compliance surveys instantly, replacing manual IT operations and saving an estimated 150 seconds per incident. • Shipped a Go-based event-driven service processing license-usage signals across 10M+ integrations spanning 360 SaaS applications, enabling auditable events with guaranteed at-least-once delivery . • Architected a multi-tenant reporting system for 1,000+ enterprise customers with scheduled execution summaries, configurable delivery preferences, and automatic catch-up logic, reducing workflow-status support inquiries by an estimated 35%. • Conducted live product demonstrations for 15 enterprise customers, translating technical architecture into business value and contributing to a JPY 300,000 paid deployment.

Software Developer Intern at Care Insurance, Jan 2025 - Jul 2025 · 7 mos, Gurugram, Haryana, India, 1) Optimized premium computation for Care Supreme by 37.5% (800ms → 300ms), supporting 1M+ daily calls by pre-computing price data at service startup instead of recalculating parameter combinations from CSV files on each request. 2) Migrated critical workflows from PHP to Node.js, improving maintainability and reducing incident risk for production backend systems. 3) Enhanced backend lead enrichment & claim workflows → fewer incidents & smoother releases

Education

Bachelor of Technology - BTech, Computer Science, National institute of technology Agartala

Higher Secondary, Udaipur English Medium Higher Secondary School

Public links

github: https://github.com/punnayan-josys