AI Engineering

MCP / AI Application Development

MCP data bridge & AI agent workflows

Overview

My AI work divides into two concrete deliverables: Advaita Data Bridge, where MCP moves data between systems through an Angular interface backed by MongoDB, and an AI-powered chatbot/agent workflow built into the Waterfall application using MCP and LangChain. Both are application features, not experiments.

Problem / Context

Data-transfer work is normally invisible — a script someone runs and hopes finished correctly. Putting it behind MCP with a real interface makes it something an operator can drive and observe.
An enterprise reporting platform holds answers that users cannot easily reach: knowing what to ask is easy, knowing which filter, cycle and region combination produces it is not. A chatbot/agent workflow is a practical route into data that already exists.
Both cases required the AI layer to integrate with an existing Angular frontend and existing backend services, rather than being built as a standalone product.

My Contribution

Advaita Data Bridge (MCP)

MCP-based development
MCP UI development
Angular integration
Backend integration
MongoDB data-transfer functionality

AI agent & chatbot workflow

LangChain development
AI-powered chatbot / AI-agent workflow
AI chatbot/agent implementation for the Waterfall application

Technologies

MCP
LangChain
Angular
Python
MongoDB

Key Engineering Areas

MCP-based development
MCP UI development
LangChain
AI agents
AI chatbots
Generative AI
Angular integration
Backend integration
MongoDB data transfer

Outcome

The data bridge is driven through an Angular MCP UI backed by MongoDB data-transfer functionality, integrated with existing backend services.
An AI-powered chatbot/agent workflow was implemented into the Waterfall application using MCP and LangChain, added to a working platform rather than replacing it.