Leaktrails
LeakTrails helps organisations strengthen their security by quickly identifying sensitive information leaks.
Speaker:
- Olivier Sagit – R&D Security Engineer at Orange Innovation
Conference Summary:
The demonstration “Demo Leaktrails” presented by Olivier Sagit addresses the issue of personal data leaks on the Internet, particularly passwords and credentials. According to estimates, around one million of these pieces of information leak daily worldwide. To counter this phenomenon, Olivier Sagit introduces Leaktrails, a cybersecurity research project focused on detecting and analyzing data leaks.
Leaktrails operates by scanning various online spaces, including the clear web, deep web, and dark web, to identify credential leaks. The goal is to qualify this data to analyze the affected sectors and targeted brands. Olivier emphasizes that the initiative aims to “trap” hackers by intervening between the moment the data has leaked and its malicious use.
Once the data is collected, the team processes and analyzes the information to extract maximum context. This allows for determining the relevance of the recovered credentials. In collaboration with partners, Leaktrails tests these passwords in their information systems, which can lead to preventive actions, such as prompting users to change their passwords before malicious exploitation occurs.
Olivier also highlights a major issue: the majority of users (65%) continue to use the same passwords for multiple accounts, exposing their digital identity to considerable risks. He concludes by reminding the importance of using different passwords for each account, despite the existence of solutions like Leaktrails.
Topics Covered:
- Personal data leaks on the Internet
- Functioning of Leaktrails
- Analysis of credential leaks
- Collaboration with partners for security
- User behaviors regarding passwords
Notable Quotes:
- “It is estimated that around one million passwords and credentials leak daily on the internet worldwide.”
- “Leaktrails is a cybersecurity research project that focuses on detecting personal data leaks.”
- “We try to position ourselves between the phase where data has already leaked and the use of these leaks.”
- “65% of people use the same password for multiple accounts, which exposes their digital identity.”
This demonstration highlights the importance of cybersecurity and best practices in password management to protect users against data leaks.
EN : Summary of the video – initial version generated by an AI on the basis of an automatic transcription and analysis of the content (Orange internal solution)
TelcoGraph
Telcograph is a platform designed to facilitate the sharing of knowledge and collaboration within the Orange group.
Yoan Chabot
AI Researcher – Project Manager
Summary:
The demonstration presents an innovation: TelcoGraph, specifically designed for a large company like Orange. Yoan Chabot explains that TelcoGraph functions as a “super Wikipedia” or an internal “super ChatGPT,” enabling employees to efficiently navigate through a vast amount of relevant data and information. The system relies on a knowledge graph that encodes information about the organization, the applications used, and general knowledge about the telecommunications and IT sectors. For example, a new employee interested in generative artificial intelligence can use TelcoGraph to discover ongoing projects, the teams involved, and the tools available within the company.
Topics Covered:
- Knowledge Graph: TelcoGraph allows for structuring and accessing internal information intuitively.
- Natural Language: Users can ask questions in natural language, making information retrieval easier.
- Response Accuracy: Unlike other language models, TelcoGraph is designed to avoid hallucinations, ensuring factual and relevant answers.
- Development Phase: Currently in the research phase, an alpha version has been deployed to gather user feedback, with the ambition to extend it to all Orange employees and potentially to other companies.
Notable Quotes:
- “In a company the size of Orange, we produce a lot of data daily, so it can be difficult to find relevant information.”
- “This tool will allow users to ask questions in natural language, which will then collect knowledge from the graph.”
- “We have built an entire technology stack around these graph solutions, which we could envision selling to other companies.”
This demonstration highlights the innovative use of artificial intelligence to improve access to information within large organizations while ensuring the reliability of the data provided.
EN : Summary of the video – initial version generated by an AI on the basis of an automatic transcription and analysis of the content (Orange internal solution)
A la Carte Home
Downloading an app has become extremely common practice for smartphone users. With this demonstration, Orange proves that it is possible to do the same with a Livebox compatible with prpl, an operating system that paves the way to new connected services. Thomas Ledru explains.
Speakers
- Jérôme Colombain: interviewer
- Thomas Ledru: Home Network Management Program Manager at Orange
Summary
Thomas Ledru presents the A la Carte Home demo, a major innovation for Livebox users: the ability to download applications directly onto their device. This advancement is made possible by prpl, a new open-source firmware developed with the contribution of Orange.
prpl allows for the installation of containers on gateways without requiring a complete update of the Livebox. However, to publish and make these services accessible via a store, a dedicated platform is necessary, which is precisely the role of A la Carte Home.
The types of applications considered include:
- Content filtering and cybersecurity services.
- Network traffic prioritization for specific uses, such as during important meetings (“quality on demand”).
- Integration of Amazon’s Greengrass environment for the connected home.
- Utilities that may be developed through an internal hackathon, “prpl Jam,” where Orange developers are invited to propose services.
User Interface: The installation process will be simplified, similar to that of application stores on iOS and Android, allowing users to select and install services with one click.
Summary of the video – initial version generated by an AI on the basis of an automatic transcription and analysis of the content (Orange internal solution)
Network AR Visualisation and Guidance
Orange has developed an application that combines augmented reality and digital mapping of the entire telecom network. With a smartphone in hand, the demonstration, with Jeremy Lacoche, puts you in the role of a maintenance technician: the surrounding infrastructure is superimposed over your field of vision.
Speakers
- Jérôme Colombain: interviewer
- Jeremy Lacoche: XR research engineer at Orange
Summary
- During the demo “Network AR Visualization and Guidance” at the Orange OpenTech event, Jeremy Lacoche presents an innovation that uses augmented reality (AR) to enhance the efficiency of field technicians. The application, named Georeso, allows for 3D visualization of Orange’s copper and fiber network mapping, thereby facilitating technicians’ interventions.
- The engineer emphasizes that one of the main challenges faced by technicians is the correspondence between 2D information and reality on the ground, particularly in locating connection points such as manholes and street cabinets. With AR, this step is simplified, reducing intervention times.
- The application also provides visual guidance by displaying procedural steps directly within the technician’s field of vision, enabling quicker resolution of customer issues. To ensure precise localization, the team uses RTK GPS technology, which allows for locating the technician with centimeter-level accuracy.
- The application is compatible with iOS and Android mobile devices, as well as tablets. In the future, Orange plans to integrate this technology into augmented reality headsets and glasses for technicians.
Quotes
- “We leverage data from the Orange information system through the Georeso application, which catalogs the mapping of the copper and fiber network.”
- “Thanks to augmented reality, we will facilitate this step, accelerate it, and avoid delays in intervention.”
- “We will guide the technician by displaying steps to follow in the technician’s field of vision.”
Summary of the video – initial version generated by an AI on the basis of an automatic transcription and analysis of the content (Orange internal solution)
Orange TV GenAI Voice Interaction
Andra Butacu demos a new way to access digital entertainment through the Orange TV service. Using your AI-powered voice control remote feature, you will be able to discover even more exciting and relevant content.
Speakers
- Jérôme Colombain (interviewer)
- Andra Butacu (developer at Orange)
Summary
This demo presents an innovation from Orange: an intelligent assistant designed to facilitate the search for television programs. Jérôme Colombain introduces the topic by highlighting the difficulties users face in finding what they want to watch on TV. Andra Butacu explains that this assistant uses advanced technologies, including artificial intelligence, to analyze user preferences and suggest suitable content. She illustrates her point with usage examples, such as the ability to search for a movie by quoting a famous line, demonstrating the assistant’s intuitive voice interaction.
The innovation relies on a rich database that allows for cross-referencing information about movies, series, and available programs, making content search faster and more efficient. This project is still in development but promises to transform the home viewing experience.
Quotes from Andra Butacu
- “This intelligent assistant is designed to find the ideal program using advanced data and algorithms.”
- “Imagine being able to say: ‘Find me the movie where the main character says I am your father.’ That’s exactly what this assistant can do.”
This presentation highlighted Orange’s advancements in voice interaction and content search, promising an enriched user experience for viewers.
Summary of the video – initial version generated by an AI on the basis of an automatic transcription and analysis of the content (Orange internal solution)
GenGraph RCA
An effective use of digital twins with Graph technology and generative AI to optimize proactive network incident analysis.
Speakers:
- Jérôme Colombain: Presenter
- Pierre-Olivier Jonery: Digital Twin Expert and Keynote Speaker
Summary:
The “GenGraph RCA” demo shows the importance of digital twins in network management, particularly during major incidents. Pierre-Olivier Jonery explains that the digital twin allows for the integration of various parts of a network, often segmented by domain (such as radio access and transmission), into a single end-to-end topology. This facilitates monitoring and data storage, enabling rapid problem analysis.
Traditionally, incident root cause analysis was a lengthy and static process, requiring manual correlation of various performance indicators (KPIs). With the use of analytical graph-based algorithms, the digital twin can isolate the root cause of a problem in seconds, whereas it previously could take up to seven hours. Furthermore, by integrating generative AI tools, it is possible to provide corrective action suggestions directly to operators.
Currently, this technology is still in the experimental phase, having been tested in a proof of concept with Orange Spain, specifically in the Pontevedra region, focusing on the radio access network and backhaul. The speakers highlight the enthusiasm among professionals at Orange for this promising tool.
Topics Covered:
- Network Failures: Challenges in identifying the causes of incidents.
- Digital Twin: An innovative tool for aggregating and analyzing network data.
- Root Cause Analysis: Transition from a static method to an automated and instantaneous approach.
- Experimentation with Orange Spain: Implementation of a proof of concept in a specific region.
Key Quotes from Pierre-Olivier Jonery:
“The digital twin will aggregate everything into a single unique graph, an end-to-end topology.”
“What used to take several hours can now be accomplished by an algorithm in a few seconds.”
“It allows for almost instantaneous understanding of the problem.” “This is an experimentation, a proof of concept that has been carried out with Orange Spain.”
This demonstration highlights technological advancements in network management and the growing importance of digital tools to enhance operational efficiency.
EN : Summary of the video – initial version generated by an AI on the basis of an automatic transcription and analysis of the content (Orange internal solution)
5G GenAI Assistant
The new generation of generative AI to simplify 5G network management.
Speakers:
- Jérôme Colombain: Journalist and Speaker 1
- Ilhem Fajjari: Infrastructure Solutions Expert
Summary:
The “5G GenAI Assistant” demo presents an innovative solution developed by Orange to automate the deployment of 5G infrastructure. Jérôme Colombain and Ilhem Fajjari discuss the virtualization of Orange’s infrastructure in the cloud, enabling virtual deployment of networks.
Ilhem Fajjari explains that the proposed tool allows administrators to make simple requests, after which the virtual assistant executes complex tasks, such as creating an infrastructure in six months within just a few minutes. This process, which previously required manual interventions and was prone to errors, is now automated while keeping humans in the loop to validate decisions.
The speakers also address concerns related to generative AI, emphasizing the importance of human expertise in assessing the accuracy of the proposals made by the assistant. The tool is designed not only for Orange administrators but also for companies needing private 5G networks, facilitating the automation and evolution of their infrastructures.
Topics Covered:
- Virtualization of Orange’s infrastructure in the cloud
- Automation of 5G infrastructure deployment
- Interaction between humans and the virtual assistant
- Evaluation of the accuracy of AI proposals
- Application of the solution for companies requiring private 5G networks
Key Quotes from Ilhem Fajjari:
“Today, we offer you an innovative solution that will automate the deployment of infrastructure and the 5G service.”
“The assistant will make proposals, and the human has the necessary expertise to validate what has been proposed.”
This demonstration highlights the importance of technological innovation in telecommunications while underscoring the crucial role of humans in the automation process.
EN : Summary of the video – initial version generated by an AI on the basis of an automatic transcription and analysis of the content (Orange internal solution)
Fiber RootCause AI Ready
An innovative AI solution to automate fiber network diagnostics on a national scale and reduce technician intervention time.
Speakers:
- Jérôme Colombain: Presenter
- Maxime Laye: AI Solutions Developer for Orange’s network
Summary:
This demo shows the integration of artificial intelligence (AI) into Orange’s network, particularly for fault detection and analysis. Maxime Laye explains that the Orange France network faces about 50 faults per day, generating up to 250,000 alarms. This volume of alerts complicates technicians’ work, requiring an average of 30 minutes to address each incident. To improve this situation, an intelligent solution has been developed to group alarms and identify the root cause of faults. The tool collects alarms in real-time and uses preventive data on network topology and associated services. The tool’s algorithms run every 15 minutes, providing instant updates via a dashboard for supervisors. Maxime Laye emphasizes that, thanks to this solution, the network remains secure, and customers experience virtually no service interruptions during faults. The solution is already operational at Orange France and in pre-production for Orange Poland.
Topics Covered:
- The impact of artificial intelligence on fault management in telecommunications networks.
- The volume and complexity of alarms generated by the Orange network.
- Reducing fault processing time through automation and data analysis.
- Network security and customer experience during incidents.
Key Quotes from Maxime Laye: “On a network like Orange France’s, there are 50 faults per day.” “Thanks to a tool, we significantly reduce this time and use AI to group these alarms.” “With a 99.99% chance, the customer sees absolutely nothing during a fault.”
This demonstration highlights the growing importance of AI in managing telecommunications infrastructure and its crucial role in improving network responsiveness and security.
EN : Summary of the video – initial version generated by an AI on the basis of an automatic transcription and analysis of the content (Orange internal solution)