Skip to content
View upadhyayanurodh's full-sized avatar

Block or report upadhyayanurodh

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
upadhyayanurodh/README.md

Anurodh Upadhyay

AI Product & Strategy — building production AI agents, one per cloud, each deployed live.

I build agentic systems at the intersection of AI product craft and enterprise infrastructure — spanning real-time operations, document-grounded RAG, and India's Digital Public Infrastructure (DPI).

What I Build

  • Real-time, event-driven systems — serverless fraud detection on AWS (Lambda · SQS · DynamoDB) with a rule engine and AI-generated risk explanations
  • Decision dashboards — single-file, zero-backend data products on S3 + CloudFront
  • Agentic apps across AWS, GCP & Azure — Amazon Bedrock (Nova), Vertex AI (Gemini 2.5 Flash + Google ADK), and Azure AI Foundry (GPT-4.1-mini), all running as live deployments
  • RAG pipelines — Azure AI Search over SharePoint document libraries, plus platform-managed RAG in Copilot Studio (Claude Sonnet 4.6)
  • Microsoft 365 / Power Platform automation — autonomous Power Automate flows, Copilot Studio agents, and Teams Adaptive Card + Outlook alerting

Current Focus

India DPI Intelligence Suite — production AI agents on India's Digital Public Infrastructure rails, each on a different cloud:

Agent DPI Rail Cloud Status
UPI Fraud Intelligence UPI (NPCI) AWS ✅ Live
ONDC Network Intelligence ONDC (Beckn) GCP 🔄 In progress
OCEN Credit Intelligence OCEN + Account Aggregator Azure 🔄 In progress

Selected Projects

  • ai-upi-fraud-intelligence-agent — Real-time UPI fraud detection on AWS. Serverless rule engine over a per-VPA sliding window, with Amazon Nova explanations and a live dashboard. Live demo
  • fmcg-ai-benchmarking-dashboard — CXO-level dashboard benchmarking enterprise AI adoption across 93 processes & 465 activities against global peers. Single file, zero backend, on S3 + CloudFront. Live demo
  • ai-gcp-pre-visit-sales-agent — Pre-visit account briefings for field reps. Gemini 2.5 Flash via Google ADK on Vertex AI, FastAPI on Compute Engine + Cloud Run. Live demo
  • ai-distributor-trade-intelligence-agent — Grounded RAG agent for field sales teams. Retrieves account documents from SharePoint via Azure AI Search, surfaces risk flags, and drafts follow-up emails. Live demo
  • copilot-studio-customer-support-agent — No-code support agent on Microsoft Copilot Studio (Claude Sonnet 4.6) with Power Automate workflow integration and platform-managed RAG.
  • ai-email-monitor-agent — Autonomous Power Automate agent that classifies every incoming Outlook email with GPT-4.1-mini and pushes Teams Adaptive Card alerts for urgent messages.
  • ai-notification-trigger-agent — Azure AI Foundry agent that reads a plain-English situation, assesses severity, and dispatches consistent alerts to Teams and Outlook via Logic Apps. Live demo

Stack

AWS GCP Azure AI Foundry Power Platform Claude Gemini Python JavaScript

Connect

LinkedIn · upadhyayanurodh@gmail.com

Pinned Loading

  1. ai-upi-fraud-intelligence-agent ai-upi-fraud-intelligence-agent Public

    Real-time UPI fraud detection agent on AWS — rule engine + Amazon Nova AI explanations + live serverless dashboard

    Python