Sudipta Paul

• Sudipta is a competent professional with 20+ years of commendable experience with IT industry in: ~Solutions Architecture ~Technical Leadership ~Onsite-Offshore Coordination and collaboration
• Experienced in Designing Cloud Native Applications, Migration of on-premise applications to public cloud, Modernization and Reengineering of Legacy Applications
• Extensive hands-on experience in architecting, solutioning and implementing large and complex projects using varied open source technologies
• Experienced in emerging technologies with varying level of expertise around Microservices, Cloud Application, Application Integration, Gen AI solutioning
• Hands-On experiance in implementing Gen AI based solution
• Architecting Computer Vission solution.
• 14+ years of experience in handling client communications from both offshore and onshore
• Worked in the following business domains - Banking, Mortgage, Travel & Hospitality, Manufacturing

• Enterprise Architecture & Solution Design
• Distributed & Cloud-Native Systems
• Microservices & API-Driven Architecture
• Digital Transformation & Modernization
• Process Optimization & Engineering Excellence

Sudipta has worked for multiple major customers who are considered world leaders in their respective domain of businesses. The range of projects that Sudipta has been a part of includes Web2.0 projects, EAI systems, Distributed system implementation,Technology Transformation projects, Cloud Migration and Cloud Native projects, Computer Vision and Generative AI project. Following are some of his noteworthy recent projects:

Project: Agentic AI Platform for Moduler and Multi-Agent Reasoning Workflows

Role: Solution Architect – Cloud & Microservices Design

Designed and developed an Agentic AI Platform enabling the definition and orchestration of autonomous agents using multiple libraries like LangChain, AutoGen, CrewAI etc. The platform supports both deterministic and non-deterministic workflows with customizable agent behaviors. Platform allows defining Tools, whether custom or built-in, and registered via a remote MCP server to ensure shared context and reusability. A custom Agent-to-Agent (A2A) communication layer was implemented for collaborative multi-agent interactions. The platform serves as a modular foundation for scalable, domain-specific AI solutions.

Key Technologies: Azure, Python 3.12, Helm Script, Azure DevOps, Docker, AKS, Terraform(IaaC), Agentic AI, Gen AI, MCP, A2A

Project: Scalable Tracking System

Role: Solution Architect – Cloud & Microservices Design

A computer vision-based application designed to monitor and track individuals on a factory floor using live camera feeds. The system detects people in real time, assigns re-identification (ReID) tags to each detected entity, and continuously tracks their movement across the floor. In multi-camera setups, it generates a unique global identifier for each person to ensure consistent identity mapping across all views. The platform supports dynamic scalability, high availability, and real-time processing, enabling accurate and uninterrupted tracking in industrial environments.

Key Technologies: AWS, AWS MSK (Managed Kafka), Python 3.12, Helm Script, Jenkins, GitLab, JFrog Artifactory, Docker, AWS EKS, Terraform(IaaC), OpenCV, RTSP, YOLO, ResNet, DeepSORT

Project: IntelliSearch – Enterprise AI-Powered Knowledge Assistant

Role: Solution Architect – Cloud & Microservices Design

IntelliSearch is an AI-driven enterprise search assistant that enables employees to interact with organizational data through natural language. Built using Retrieval-Augmented Generation (RAG), it supports use cases like policy lookup, document discovery, and process guidance. The solution uses Azure AI Search as a vector store and LLaMA 3 as the foundation model. It is deployed on Azure AKS for scalable inference and integrates event-driven ingestion pipelines for real-time data indexing. The system is designed for high availability, secure access, and enterprise-grade scalability.

Project: Next Gen (*name changed),  Internet-Facing Multi-Cloud Customer Portal
Role: Solution Architect – Cloud & Microservices Design

Next Gen was a cloud-native, internet-facing application portal, re-architected to operate across a multi-cloud environment integrating AWS, Adobe AEM, OKTA, Google Apigee, and Confluent Kafka. The platform was built on a microservices architecture using Domain-Driven Design and followed an API-first development approach to ensure modularity, scalability, and performance. It supported federated authentication, real-time data streaming, and dynamic content management, delivering a seamless and secure user experience. The solution was deployed with high availability and resilience across cloud environments, enabling future-ready digital transformation.

Key Technologies:
AWS, Google Apigee, OKTA (OAuth2), Adobe AEM, Confluent Kafka, Spring Boot (Microservices), Docker, AWS ECS, App Mesh (service mesh), OAuth2, CI/CD, CloudFormation Template (IaaC), Datadog