FinBot delivers 24/7 customer support for European insurance companies via WhatsApp — contract Q&A, appointment booking, and claims filing. GDPR compliant, deployed on AWS Frankfurt, built by Saurabh Mukherjee.
I'm a freelance AWS Cloud Architect and AI Engineer specialising in Generative AI solutions for European financial and insurance companies. My focus is on building production-grade systems that are secure, GDPR-compliant, and cost-efficient.
I work with Amazon Bedrock, Lambda, DynamoDB, and the full AWS ecosystem to deliver AI assistants that integrate with your existing systems — without requiring access to your database.
Multi-account AWS Organizations, SCPs, IAM Identity Center, Control Tower roadmap.
RAG pipelines, Knowledge Bases, Bedrock model integration, prompt engineering.
EU data residency (Frankfurt), data processing agreements, audit-ready architecture.
End-to-end WhatsApp bots — contract Q&A, appointment booking, live financial data.
A fully deployed WhatsApp bot that answers questions from your documents, connects to your financial systems, and books appointments — 24/7, in German and English.
Production-grade AWS infrastructure designed for European regulated industries — secure, auditable, and cost-efficient.
From idea to working demo in weeks. I build proof-of-concepts that you can show to decision makers — not slide decks.
End-to-end AI assistant for a Bavarian insurance company. Customers interact via WhatsApp to get answers from insurance contracts, book appointments, and receive email confirmations — all in German. Deployed on AWS Frankfurt for GDPR compliance with a real-time Amplify dashboard.
Full cloud migration from on-premises infrastructure to AWS with GenAI capabilities enabled post-migration. Includes secure containerisation, CI/CD pipelines, and German language support on Amazon Bedrock.
AI-powered CA (Chartered Accountant) assistant delivered via WhatsApp. Uses RAG pipeline on Amazon Bedrock to answer Indian tax, GST, and TDS queries with Knowledge Base retrieval and DynamoDB conversation history.
Enterprise legal research agent built on Amazon Q Business. Enables lawyers to search across thousands of legal documents using natural language, with source citations and Google Drive integration.
Production-grade MLOps pipeline deploying an LSTM deep learning model to ECS Fargate. Includes automated retraining, model versioning, GitHub Actions CI/CD, and CloudWatch monitoring.
Customer photographs a damage claim via WhatsApp. AI extracts fields using Amazon Textract, pre-fills the insurance form, and queues for human review — no website, no login required.
Available on Upwork for quick-start projects with defined scope and delivery timelines.
AI readiness assessment and architecture roadmap for your business.
View on UpworkSecurity gaps, cost waste analysis, and top 5 actionable fixes.
View on UpworkFull WhatsApp AI assistant built on AWS Bedrock with RAG — production ready.
View on Upwork