tl;dr
For MWC 2025, we built an AI-powered demo that pulls real-time data from our CSRD system—aggregating network monitoring metrics, Jira tickets, and SLA stats—to provide customers with clear, contextual answers about network incidents. Inspired by Perplexity AI, we designed the assistant to not only summarize incidents but also back up every answer with interactive data widgets. The MVP struck a balance between technical feasibility, sales needs, and UX best practices, resulting in strong interest from major telecom clients at the event.
>
Concept: From Data Chaos to Customer-Centric Insights
The Problem
Telecom clients often struggle to understand how network incidents impact their business. Standard chatbots without real data references feel unreliable and generic.
Solution:
An AI assistant that:
Analyzes data from the CSRD system (Jira tickets, TTR, network stability metrics).
Generates contextualized answers, e.g., “A 40% speed drop in area X on 29.05.2025 was caused by equipment updates. 12 corporate customers were affected, all have been notified.”
Supports answers with widgets: SLA graphs, affected customer lists, ticket statuses.
Solution Design
UX Principles:
3-second rule: First screen—concise answer with key metrics (e.g., “SLA: 99.3%”).
Depth on demand: Ability to expand sources (charts, customer lists) via tabs.
Business language: Swapped technical jargon for business-friendly phrases like “Your client Y was not impacted because…”.
Sample Query:
“Were there any outages in Madrid on May 28?”
→ Response:
Summary: “Yes, 2 incidents. Average recovery time—1h 20min.”
Sources: Outage map, list of 8 affected customers, Jira ticket link.
>
Outcome & Reflection
At MWC 2025, the AI-powered demo impressed both clients and sales teams by delivering clear, conversational analytics, turning complex network data into intuitive insights and visualisations, such as answering “Why did my VoIP lag?” with real-time network load charts.
Thanks for checking!
>
More projects
