Sanjay Chintamani Patel (Keenpreneur)
Full Stack AI Architect
We denounce with righteous building enterprise AI and cloud-native solutions. Specialized in Generative AI, Agentic AI, RAG, and Multi-Agent Systems using Azure, AWS, LangChain, & FastAPI.
Years Of Experience
Project Complete
Global Accreditation
Professional Problem Solutions For Digital Products
I am a passionate Full Stack AI Architect & Engineer with over 7+ years of experience building scalable, enterprise-grade AI platforms and cloud-native applications across BFSI, Healthcare, ERP, Retail, and Data Intelligence domains. My expertise lies in Generative AI, Agentic AI, Multi-Agent Systems, RAG architectures, and cloud engineering, where I combine intelligent automation with modern software engineering to deliver secure, high-performance AI ecosystems.
- Agentic AI Development
- AI Agent Development
- Multi-Step AI Agent LLM
- Advanced RAG (Hybrid Vector)
- AI Orchestrator with A2A/MCP
Total Work Experience
Full Stack AI Architect
Tata Consultancy ServicessMLOps Engineer
Mobile Programming LLCAI ML Engineer
iAapTeck Software LabsSenior AI/ML Analyst
CapgeminiSoftware Developer Internship (Python Flask Development)
Myospaz Software TechnologiesSkill Set & Expertise
LLMS & Foundation Models
GPT-40, Claude (Anthropic), Gemini, Llama 2/3, Mistral, Mixtral, DeepSeek, Phi-3/Phi-4, Ollama, OpenAI APIs, & Hugging Face Transformers
Al Solution Capabilities
Al Agent Development, Multi-Agent Orchestration, Enterprise RAG Systems, Conversational Al, Copilot Development, LLM Fine-Tuning, and Vector Search
Agentic Al Frameworks
Lang Chain, LangGraph, LlamaIndex, CrewAl, AutoGen, Semantic Kernel, Langflow, LangFuse, LangWatch, Haystack, DSPY, OpenDevin, RAG, AI Agents, Multi-Agent Systems, Prompt Engineering, Function Calling, & MCP
Cloud & Al Platforms
Microsoft Azure (Azure OpenAI, Al Foundry, Al Search, Azure ML, Functions, Logic Apps, API Management, AKS, Key Vault), AWS (Bedrock, SageMaker, EC2, ECS, EKS, Lambda, S3, CloudWatch), GCP (Vertex Al, Gemini APIs, BigQuery, Cloud Functions, GKE, Cloud Run)
AI/ML Libraries & NLP
PyTorch, TensorFlow, Scikit-learn, Pandas, NumPy, XGBoost, spaCy, NLTK, OCR, Computer Vision, NLP, ML Algorithms, and Random Forest
AI/ML Libraries & NLP
PyTorch, TensorFlow, Scikit-learn, Pandas, NumPy, XGBoost, spaCy, NLTK, OCR, Computer Vision, NLP, ML Algorithms, and Random Forest
DevOps, MLOps & Observability
Docker, Kubernetes, MLflow, Kubeflow, GitHub Actions, CI/CD Pipelines, Azure DevOps, Terraform, Splunk, Dynatrace, Prometheus, and Grafana
Storage & Databases
Azure Cosmos DB, Azure Blob Storage, Azure SQL Database, Azure Data Factory, Databricks, MySQL, MongoDB, DynamoDB, Firestore, AlloyDB, and RDS
Programming
Python, FastAPI, REST APIs, C#.NET, ASP.NET Core, JavaScript, TypeScript, Node.js, SQL, Bash/Shell Scripting, and YAML
Others
Android Studio, REST API, RETROFIT, SQLLite, ASP.Net Core Enitity Framework, LINQ, C#.Net, PHP
Projects & Client Works Indeed
TCS: DNB e-Zenit AI Customer Onboarding & Intelligent Middleware
Led the architectural redesign of a mission-critical banking onboarding platform into a fully Agentic AI ecosystem. Built multi-agent onboarding workflows using LangChain and LangGraph for automated KYC/AML compliance, engineered enterprise-grade RAG pipelines via Azure Cosmos DB Vector Search, and established secure AI middleware to orchestrate open-source and private LLMs.
TCS: DNB CENTRIX - Centralized Incident Tracking & Execution Platform
Designed and developed an AI Agent-enabled incident management platform using Hybrid RAG architecture (vector + non-vector). Integrated knowledge bases like Confluence and code repositories to enable specialized triage agents to analyze logs, provide step-by-step debugging recommendations, and route tickets autonomously.
TCS: Ultimatix Gigs Enterprise Solution - AI Enabled Chatbot Portal
Built an enterprise-wide multi-agent chatbot portal automating core HR and IT service workflows. Utilized CrewAI, LangGraph, and Model Context Protocol (MCP) to coordinate domain-specific agents (HR, Policy, Timesheets) linked via secure RAG pipelines to backend enterprise APIs
Programming LLC: SISCore DTSS ERP - AI Driven ERP Automation & Predictive Analytics Platform
Functioned as an MLOps Engineer to inject intelligent automation into traditional ERP workflows. Developed predictive machine learning models for sales forecasting and payment prediction using XGBoost/Scikit-learn, built NLP/OCR document parsing pipelines, managed containerized cloud deployments on Azure AKS.
iAapTeck: BolBhidu CPRS Digital Solution - AI Driven Content Personalization & Recommendations Platform
Built low-latency content recommendation engines utilizing collaborative filtering, K-Means, and DBSCAN clustering algorithms. Orchestrated high-traffic user behavioral analytics and implemented automated MLOps pipelines using Azure Machine Learning and Docker for continuous deployment
iAapTeck: Kidebaj AI-SRP Smart Retail Platform
Designed an AI-powered e-commerce ecosystem incorporating personalized product discoveries, semantic search, and inventory forecasting. Integrated Azure Cognitive Services and Azure AI Vision for automated image analysis, product tagging, and context-aware marketing chatbots.
Capgemini: Equinor PAIDPS Portal - AI Driven Predictive Analytics & Intelligent Document Processing System
Designed an AI-powered e-commerce ecosystem incorporating personalized product discoveries, semantic search, and inventory forecasting. Integrated Azure Cognitive Services and Azure AI Vision for automated image analysis, product tagging, and context-aware marketing chatbots.
Capgemini: Equinor Statoil
I partnered with Equinor to create a two-way orchestration portal via FASTAPI, optimizing supplier-customer interactions. Leveraging Azure for data storage, computation, and containerized code, the portal improved data flow and operational efficiency, elevating Equinor’s energy solutions.
iAapTeck: iLearning Career Platform
Led backend development for the iLearning Career Platform, utilizing Python with FastAPI. Integrated with Angular for a seamless user experience, ensuring efficient data handling and responsive design to enhance the educational journey for users.
iAapTeck: Tivhi Play App
Developed Tivhi Play App, an AI-enabled content engagement platform using Java, XML, PHP, Python, FastAPI & phpMyAdmin on Android Studio. Designed UI/UX in Figma and integrated AI features for tailored recommendations, delivering a seamless user experience with interactive engagement rewards.
Conferences & Events
Skill Set & Expertise
Explore My Popular Projects Repository
Fast Bottom Navigation Bar Library
FastBottomNavigationBar is a lightweight and fast Android library designed to implement Material Design bottom navigation bars. Written entirely in Kotlin, it integrates seamlessly with Android Jetpack's Navigation Component to automate fragment transactions by aligning menu item IDs with your navigation graph. It is highly customizable, allowing developers to easily adjust XML styling properties such as indicator colors, corners, badges, and icon tints right out of the box.
MongoDB Atlas RAG Ingestion
MongoDB-Atlas-RAG-Ingestion is a Python-based, end-to-end ingestion pipeline designed for building Retrieval-Augmented Generation (RAG) applications. It leverages LangChain to chunk source PDF data, OpenAI to extract intelligent metadata, and VoyageAI to generate high-quality vector embeddings. The processed documents are stored directly in MongoDB Atlas, enabling semantic similarity searches to fetch context and feed a generative LLM for accurate, context-aware answers.
GCP-to-AWS Agentic Transpiler
GCP-to-AWS-Agentic-Transpiler is a Python-based application built with LangGraph that automates the conversion of GCP Terraform scripts into AWS equivalents. Driven by a 6-agent pipeline, it recursively scans directories, maps dependencies across 25+ major services, and generates fully configured AWS output files. The framework features cross-provider LLM support, allowing seamless integration with OpenAI, Claude, Gemini, and local Ollama deployments.
Smart Real-Estate RAG System
Smart-Real-Estate-RAG-System is a Python-based real estate assistant that combines a Neo4j Knowledge Graph with ChromaDB vector search to handle structured, semantic, and hybrid queries. Driven by LangChain and OpenAI, it uses a smart routing pipeline to classify user intent, dispatching requests to either a Cypher-generating graph engine or a self-healing Corrective RAG workflow. Served via Streamlit, the system includes automatic fallback mechanisms to seamlessly merge multi-source data or switch engines when confidence is low.
Detect Text Image Moderation - Azure AI Content Safety Studio
AzureAI-Content-Safety-Studio--Detect-Text-Image-Moderation is a Python-based repository designed to moderate user- and AI-generated text and images using Azure AI Content Safety. Deployed via Azure Container Apps and integrated with Foundry Agent Service, it provides scripts to detect harmful material and user input attacks alongside an interactive Content Safety Studio to test and sample moderation APIs.
HITL AI Workflows Azure Agent
HITL-AI-Workflows-Azure-Agent is a Python-based repository demonstrating the Contoso Fraud Detection & Response Workflow using the Azure Agent Framework. It implements a Human-in-the-Loop (HITL) architecture where multi-agent systems leverage Azure OpenAI to analyze suspicious activities and automatically route high-risk events to a real-time React and FastAPI dashboard for human analyst review.
Azure RAG Chat Python - A RAG Chat with Azure-AI Search OpenAI & Python
Azure-RAG-Chat-Python is a Python-based sample application that implements Retrieval-Augmented Generation (RAG) using Azure AI Search and Azure OpenAI Service to build a ChatGPT-like experience over custom documents. It features a multi-turn chat interface with citations and reasoning steps, optional multimodal support for image-heavy data, and secure login via Microsoft Entra integration. The repository is designed for rapid deployment using the Azure Developer CLI and supports hosting on Azure Container Apps.
Enterprise Agent by Foundry AI Knowledge
Enterprise-Agent-by-Foundry-AI-knowledge is a Python-based repository designed to build an AI-powered ticket management system using the Model Context Protocol (MCP). It features an Azure Search MCP tool to query protected knowledge and connects multi-agent workflows to external APIs like Freshdesk, automating ticket creation and request routing based on complex user inputs.
Building Multimodal AI Applications Using GPT-4o with Azure OpenAI
Building-Multimodal-AI-Applications-Using-GPT-4o-with-Azure-OpenAI is a Python-based repository demonstrating an AI-powered workflow for video and multimodal content generation using Azure OpenAI. It provides hands-on configurations and Jupyter notebooks to deploy and manage GPT-4o models within Azure AI Foundry, enabling developers to analyze visual inputs and integrate multimodal intelligence into web applications.
Let’s Talk For your Next AI Projects
Let’s transform your business with intelligent automation and modern software engineering. Partner with a seasoned Full Stack AI Architect to build scalable, secure, and enterprise-grade AI platforms tailored to your business domain.
- 7+ Years Of Experience
- Generative & Agentic AI Solutions
- Cloud-Native Architecture
- Custom RAG & Multi-Agent Systems