BARQ – AI-Powered ServiceNow Support Assistant | AI Agents & RAG

**Project Overview**

As part of the BARQ Systems × Sprints AI Hybrid Internship, I contributed to developing an AI-powered ServiceNow Support Assistant designed to help IT support teams analyze incidents, retrieve relevant technical knowledge, and generate intelligent resolution suggestions.

The system integrates ServiceNow with an intelligent backend powered by AI, semantic search, and Retrieval-Augmented Generation (RAG), while keeping human review at the center of the resolution process.

**Key Features & Technical Implementation**

**1. ServiceNow Integration**

- Configured asynchronous Business Rules to capture incident events.

- Integrated ServiceNow with an external backend using REST APIs.

- Implemented automated incident event transmission.

**2. Secure API Communication**

- Implemented HMAC-SHA256 request signing.

- Validated incoming webhook requests to ensure payload integrity and secure communication.

**3. Backend Development**

- Developed a FastAPI webhook for receiving and processing incident events.

- Used Redis for event deduplication.

- Integrated Celery for asynchronous background task processing.

**4. AI & Semantic Search**

- Worked with text embeddings to represent technical knowledge.

- Integrated Qdrant Vector Database for semantic similarity search.

- Retrieved relevant knowledge to provide context-aware assistance.

**5. AI Agent & RAG**

- Integrated an AI-powered agent with retrieved knowledge.

- Generated suggested incident resolutions based on relevant information.

- Maintained human oversight throughout the support workflow.

**Tools & Technologies**

- Python

- FastAPI

- ServiceNow

- REST APIs

- AI Agents

- Retrieval-Augmented Generation (RAG)

- Qdrant

- Vector Embeddings

- Redis

- Celery

- HMAC-SHA256

- Backend Integration

**Project Outcome**

Contributed to building an integrated AI-powered IT support solution that connects enterprise incident management with intelligent knowledge retrieval and automated assistance.

This project provided practical experience in developing secure backend integrations, asynchronous processing pipelines, AI agents, and RAG-based applications for real-world business scenarios.