Pankaj Parihar
Confidential

WhatsApp CRM & AI Automation Platform

Enterprise CRM/ERP modules with end-to-end WhatsApp Cloud API messaging and an AI chatbot on Azure OpenAI + Azure AI Search.

Role
Software Engineer — Angular front-end, ASP.NET Core APIs, integrations
Timeline
Nov 2024 – Present
Organization
Dial ERP Technology
Status
Confidential — screenshots only
Illustrative UI mockup of the CRM inbox — WhatsApp conversations, a template message and AI-suggested replies.

Illustrative UI mockup, not a real screenshot — real screens can be shown in an interview. Built at Dial ERP Technology; no proprietary code or client data shown.

TL;DR

  • 40+ REST APIs in ASP.NET Core
  • WhatsApp templates, media, webhooks & bulk messaging
  • AI chatbot on Azure OpenAI + Azure AI Search (RAG)
  • Reusable Angular Material component library

Overview

At Dial ERP Technology I work on an enterprise CRM & ERP platform built with Angular, ASP.NET Core Web API (C#) and MongoDB. Two parts of my work stand out: end-to-end WhatsApp messaging through Meta’s WhatsApp Cloud API, and AI chatbot features powered by Azure OpenAI and Azure AI Search.

My role

  • Built CRM & ERP modules across the stack — Angular (Material) front-ends and ASP.NET Core Web APIs on MongoDB.
  • Designed and shipped 40+ REST APIs and a set of reusable Angular Material components.
  • Integrated the WhatsApp Cloud API: template messages, media, webhooks and bulk messaging.
  • Implemented AI chatbot features with Azure OpenAI and Azure AI Search.
  • Production support, bug fixing and performance optimization; Agile sprints with code reviews.

What the WhatsApp integration covers

  • Template messages — WhatsApp requires pre-approved templates to start a conversation, so the CRM sends approved templates with dynamic values.
  • Media — images and documents sent and received inside the CRM.
  • Webhooks — Meta calls the API back with delivery and read status and with incoming replies, so agents see conversations update in the CRM.
  • Bulk messaging — campaigns to many contacts from one screen.

How the AI assistant answers

The chatbot uses retrieval-augmented generation (RAG): Azure AI Search finds the most relevant company data for a question, and Azure OpenAI writes an answer grounded in those results — so replies are based on the business’s own data instead of guesswork.

Architecture

WhatsApp message flow The Angular CRM sends messages through the ASP.NET Core Web API to Meta's WhatsApp Cloud API, which delivers them to the customer. Delivery status and replies come back to the API through webhooks, are stored in MongoDB, and update the CRM. CRM web app Angular · Material ASP.NET Core API C# · REST + webhooks WhatsApp Cloud API Meta Customer on WhatsApp MongoDB messages · status send / bulk template · media deliver reply webhook: status, replies live updates store / read
Simplified message flow — outgoing messages (solid) and webhook callbacks (dashed). Details anonymized.
AI assistant answer flow (RAG) A question from the CRM chat goes to the ASP.NET Core API. The API retrieves relevant company data from Azure AI Search, sends the question with that context to Azure OpenAI, and returns a grounded answer to the CRM. Question CRM chat · Angular ASP.NET Core API orchestrates the flow Azure AI Search finds relevant company data Azure OpenAI writes a grounded answer ask ① retrieve ② question + context ③ answer
Simplified RAG flow — retrieval first, then a grounded answer. Details anonymized.

Tech stack

  • Angular
  • Angular Material
  • TypeScript
  • ASP.NET Core Web API
  • C#
  • MongoDB
  • Azure OpenAI
  • Azure AI Search
  • WhatsApp Cloud API

Ask Pankaj

Scripted assistant · instant answers

Hi 👋 I’m Pankaj’s assistant. Ask about his experience, skills or availability — or tap a question below.

30-second summary

Pankaj Parihar

Software Engineer · Angular · ASP.NET Core · Azure OpenAI

Open to new opportunities

Experience
2+ years
Current role
Software Engineer, Dial ERP Technology
Location
India · IST
Education
MCA, MBM University (84%)

Core stack

Highlights