About Skills Experience

Agentic AI & Backend Engineer

I build agentic AI platforms and scalable backend systems at the intersection of LLM orchestration and cloud engineering. Currently shipping production AI at Epsilon–Publicis Groupe.

0+

Years Exp.

0%

Faster Insights

0+

API Tools Built

Kaustav Paul — AI Software Engineer 2
LangGraph
Spring Boot
Agentic AI
Java
LLM
GCP
scroll
LangGraph LangChain Java / Spring Boot TypeScript Node.js AWS Bedrock Google Gemini GCP Cloud Run Terraform PostgreSQL MCP Docker LangGraph LangChain Java / Spring Boot TypeScript Node.js AWS Bedrock Google Gemini GCP Cloud Run Terraform PostgreSQL MCP Docker

Building AI systems
that actually ship

📍 Bangalore, IN
Open for relocation

Hey! I’m Kaustav Paul — a Software Engineer 2 with 4 years of experience building scalable backend systems and agentic AI platforms in the AdTech and retail-media domains.

I led a team of 5 engineers to architect and deliver a production LangGraph-based conversational AI platform integrating MCP, multi-provider LLM orchestration and real-time SSE streaming. I have a strong foundation in Java/Spring Boot, REST API design, and cloud infrastructure on GCP.

Outside of work, I tinker with hardware — I built Neo, an open-source voice-controlled AI robot on Raspberry Pi 5 powered by Gemma3 with on-device wake-word detection and an agentic tool-calling architecture.

🧠

Agentic AI

Production LangGraph/LangChain ReAct agents with MCP tool integration and multi-provider LLM orchestration.

Backend Engineering

Java/Spring Boot microservices, REST APIs, Hibernate & PostgreSQL for high-scale partner integrations.

Cloud Infrastructure

GCP Cloud Run, Cloud SQL, pgvector — provisioned reproducibly with Terraform IaC.

🔧

Hardware Hacker

Raspberry Pi robotics, embedded Python, I2C sensors, servo control — bridging AI with the physical world.

My technical arsenal

AI & Agentic Systems

LangGraph
LangChain
Model Context Protocol
ReAct Architecture
Prompt Engineering
LLM Orchestration
RAG
Promptfoo

Backend Engineering

Java
Spring Boot
TypeScript
Node.js
Python
RESTful APIs
Microservices
Server-Sent Events
Hibernate
PostgreSQL

Cloud & Infrastructure

GCP Cloud Run
Cloud SQL
pgvector
Terraform (IaC)
Docker
Kubernetes
CI/CD Pipelines
AWS Bedrock

Proficiency

Agentic AI / LLM Systems92%
Backend Engineering (Java / Node)90%
Cloud & Infrastructure (GCP / Terraform)82%
System Design & Architecture85%

My journey

Software Engineer 2

Epsilon — Publicis Groupe

Jan 2023 — Present

Led a cross-functional team of 5 engineers to deliver a production agentic AI platform in 4 months — 33% faster than the team’s typical 6-month cycle. Reduced campaign analysis time from 20 min to 2 min (90% reduction). Designed a LangGraph ReAct agent backend with SSE streaming, built an MCP server with 80+ typed tools, and implemented multi-provider LLM orchestration (AWS Bedrock/Claude, Gemini, Ollama). All services provisioned on GCP via Terraform.

LangGraphMCPJava/Spring BootTypeScriptGCPTerraform

Master’s in Computer Application

RV College of Engineering, Bangalore

Jan 2021 — Jan 2023

Postgraduate specialisation in advanced computing, software engineering principles, distributed systems, and applied machine learning. Graduated from one of Bangalore’s top engineering institutions.

Distributed SystemsMachine LearningSoftware Engineering

Bachelor’s in Computer Applications

Acharya Institute of Graduate Studies, Bangalore

Jan 2017 — Jan 2020

Foundation in computer science fundamentals, programming languages, database management, and software development. Built the base for a career in full-stack and systems engineering.

Computer ScienceDatabase SystemsProgramming

Let’s build something
incredible together

Have a project in mind, an open role, or just want to say hi? I’m always happy to chat — drop me a message and I’ll get back to you within 24 hours.