AI Agent Examples

1. Telegram RAG Workflow

Image: Telegram RAG Workflow Overview

The Telegram RAG workflow integrates Telegram messaging with document ingestion, vector search, and AI-driven responses.

The Telegram RAG demo creates a fully automated loop between Telegram and Omniscope, turning uploaded documents into an interactive knowledge source.

Document ingestion and vector storage

• When Telegram users upload PDFs, the workflow automatically fetches and processes them.

• The text is extracted, chunked, and stored in a ChromaDB vector database for semantic search.

• Telegram is notified once the documents have been processed.

Retrieval-Augmented Question Answering

• User questions arriving from Telegram are matched against the stored embeddings.

• Relevant PDF excerpts are fed to an AI Completion block to produce accurate, context-aware answers.

• The enriched answers are sent straight back to the user on Telegram.

This workflow demonstrates how Omniscope can serve as a RAG-powered backend for messaging platforms, enabling document-aware conversational support.

2. Google Analytics Weekly Report

Image: Google Analytics Weekly Reporting Workflow
The Google Analytics Weekly Report workflow automates data retrieval, AI-generated summarisation, and multichannel delivery.

The Google Analytics Weekly Report demo automates the production of a weekly sales summary based on Google Analytics 4 data.

The workflow:

• Retrieves sales data for the latest 7 days as well as the equivalent week one year earlier.

• Aggregates and compares both time periods to reveal trends and changes in performance.

• Passes the combined results to AI Completion blocks to create:

  • a detailed, email-ready narrative report

  • and a concise, Telegram-friendly summary

• Automatically sends the formatted outputs via email and Telegram.

This demo highlights how AI can streamline routine reporting processes—transforming raw analytics into human-readable insights and distributing them effortlessly.