Headline

Sr SDE @ Oracle | Backend & Distributed Systems | GenAI / Agentic AI / LLM Platforms | Java, Go, Spring Boot, Kafka | Ex-Target

About

Backend software engineer with ~10 years designing and building distributed systems, APIs, and GenAI infrastructure across Oracle, Target, Carelon, TechMahindra, and Ericsson. Currently at Oracle as Principal Member of Technical Staff (Staff SWE equivalent), contributing to a cloud-native data platform on Java, Spring Boot, Kafka, Kubernetes — and an agentic AI engine on LangGraph + Java. ━━━━━━━━━━━━━━━━━━ 🔧 BACKEND SYSTEMS & APIs I'VE BUILT ▸ GenAI Orchestration Platform (Target) — REST + gRPC APIs for a multi-tenant LLM service. Async request handling, provider abstraction across OpenAI / Anthropic / Llama, per-tenant rate limiting, circuit-breakers, retry with backoff, Redis caching. Adopted by 15+ teams (60% integration growth). ▸ Document Extraction Service (Target) — Java 17 / Spring Boot REST API + Kafka exactly-once consumer pipeline. Async job submission, status polling + webhooks, S3 blobs, PostgreSQL metadata, multi-tenant isolation. 8K+ jobs/day. Saved 1,200+ hrs/month. ▸ Claims Microservices (Carelon) — 20+ Java/Spring Boot services, Kafka event sourcing on AWS, sync REST + async consumers. JFR / async-profiler, G1GC tuning. 30% latency cut, 25% P99 gain, 99.97% availability. ▸ API Gateway (TechMahindra) — Apigee Edge fronting 30+ partner APIs. OAuth 2.0, rate limiting, request transforms, traffic shadowing. 500K+ req/hour. ━━━━━━━━━━━━━━━━━━ 🤖 GENAI / AGENTIC AI INFRASTRUCTURE ▸ Multi-provider LLM routing with pluggable adapters, token-aware load balancing, fallback chains, streaming. ▸ Llama fine-tuning with LoRA adapters for domain extraction. 62% → 87% accuracy. ▸ Multi-agent system on LangGraph + Java (Oracle) — multi-step reasoning, tool-calling agents, RAG over OCI GenAI vector store. ━━━━━━━━━━━━━━━━━━ What I'm known for: turning vague requirements into clean API contracts and event-driven architectures, shipping to production, and instrumenting them so well that on-call stays quiet. ━━━━━━━━━━━━━━━━━━ ⚙️ TECH STACK Languages — Java (primary), Python, Go, C++, SQL Backend & APIs — Spring Boot, REST, gRPC, GraphQL, Apigee, OAuth 2.0, JWT, OpenAPI Distributed Systems — Kafka (exactly-once), Kubernetes, Docker, Helm, ZooKeeper, OpenTelemetry, JFR / async-profiler Cloud & Storage — AWS (EC2, S3, Lambda, SQS, EKS, RDS), OCI, GCP, PostgreSQL, Redis, Elasticsearch, DynamoDB GenAI — LangGraph, LangChain, RAG, OpenAI / Anthropic, Llama, LoRA, OCI ━━━━━━━━━━━━━━━━━━ Open to Senior SWE/ Staff IC roles in backend, distributed systems, or GenAI infrastructure. 📩 kumar.shubham71994@gmail.com

Location

Bengaluru

AI stack

claude

codex

github copilot

hugging face

langchain

langgraph

Work experience

Senior Software Engineer at Microsoft, Jul 2026 - Present · 1 mo, Bengaluru

Principal Member of Technical Staff at Oracle, Feb 2026 - Present · 5 mos, Greater Bengaluru Area, Backend & platform engineer on a cloud-native data platform. Contributing to high-throughput data ingestion services on Java, Spring Boot, Kafka, Kubernetes — and to a multi-agent system on LangGraph + Java with tool-calling agents and RAG over OCI GenAI.

Senior Software Engineer at Target, Sep 2024 - Jan 2026 · 1 yr 5 mos, Bengaluru, Karnataka, India, ▸ Designed and shipped a multi-tenant GenAI orchestration platform from scratch. REST + gRPC APIs, async multi-provider LLM routing (OpenAI / Anthropic / Llama), circuit-breakers, token-aware load balancing. Adopted by 15+ teams (60% integration growth). ▸ Built a Kafka exactly-once document extraction pipeline (Java 17 / Spring Boot, S3, PostgreSQL, multi-tenant) processing 8K+ jobs/day. Saved 1,200+ person-hours/month. ▸ Fine-tuned Llama with LoRA adapters; accuracy 62% → 87%.

Senior Software Engineer at Carelon Global Solutions, Dec 2019 - Sep 2024 · 4 yrs 10 mos, Bengaluru, Karnataka, India, ▸ Designed and built 20+ Java/Spring Boot microservices with Kafka event sourcing on AWS. JFR / async-profiler tuning, G1GC optimization. 30% latency cut, 25% P99 improvement, 99.97% availability. ▸ Led monolith-to-Kubernetes migration (Docker, Helm, HPA). Cut release cycles from 4 weeks to 2.

Software Engineer at Tech Mahindra, Mar 2017 - Nov 2019 · 2 yrs 9 mos, Bengaluru, Karnataka, India, Designed an Apigee API gateway with OAuth 2.0 and scalable patterns handling 500K+ req/hour. Re-engineered a 6-hour PL/SQL billing job into Spring Boot microservices completing in 40 minutes. Rockstar Award.

Associate Software Engineer at Ericsson, Aug 2016 - Feb 2017 · 7 mos, Bengaluru, Karnataka, India, Built 6 Spring MVC REST APIs for the Network Performance Platform with 80%+ JUnit/Mockito coverage. Exceeded 97% availability KPI.

Internship at Ericsson, Feb 2016 - Jul 2016 · 6 mos, Bengaluru, Karnataka, India, Maintained KPI dashboards and resolved data discrepancies for site performance monitoring.

Education

Bachelor of Engineering - BE, Electrical, Electronics and Communications Engineering, Kalinga Institute of Industrial Technology, Bhubaneswar