Internship: Autonomous AI Agents with RAG Framework for Dynamic Network Management

Internship: Autonomous AI Agents with RAG Framework for Dynamic Network Management
Contacts: Mehdi Hmidi mehdi.a.hmidi@ericsson.com, Fredrik Öström
fredrik.ostrom@ericsson.com
Final Year Engineering School or Master 2 Level
Duration: 6 months
Starting Date: February 2025
Location: Ericsson R&D center Massy- France

Project Overview

Ericsson’s R&D center in [Location] is at the forefront of AI-driven innovation in 5G/6G and cloud computing. As part of our Standards & Technology unit, we are seeking an intern to contribute to our research in applying Generative AI models and develop autonomous AI systems using Retrieval Augmented Generation (RAG) frameworks and multi-agent orchestration. This project draws inspiration from pioneering AI thought leaders like Marvin Minsky, John McCarthy, and Patrick Winston, focusing on creating intelligent agents capable of advanced reasoning and dynamic decision-making for network optimization.

Project Description

The intern will explore the development of an autonomous multi-agent system that leverages RAG for knowledge retrieval and real-time adaptation to complex network
conditions. The project emphasizes the following:

• Dynamic Agent Deployment: Implement a system that can spin up AI agents on demand, creating specialized agents to tackle specific network challenges such as a resource allocation, anomaly detection, or latency management. These agents will be activated dynamically based on the network’s real-time needs.
• On-the-Fly Delegation: Develop mechanisms for agent-to-agent delegation, allowing primary agents to assign tasks to newly created or existing agents. This delegation strategy will enable efficient handling of network events, such asoverload situations or unexpected traffic spikes.
• Plan Restructuring and Adaptation: Create a plan restructuring capability that allows the multi-agent system to adapt its strategies based on changing conditions.This involves dynamic re-planning of resource allocations, user association strategies, and network recovery processes when faced with evolving scenarios.
• Higher Reasoning Layer Development: Design and integrate a higher reasoning layer where agents collaborate to synthesize information from various sources, conduct complex analysis, and make strategic decisions. This layer will draw on concepts from Minsky's Society of Mind, focusing on combining reflexive, symbolic, and deliberative reasoning to improve decision quality.
• Integration with RAG Framework: Utilize the RAG framework to retrieve knowledge from structured databases, unstructured logs, and domain-specific knowledge bases, feeding relevant information into the reasoning layer to enhance decision-making accuracy.
• Simulation and Validation: Use Ericsson’s digital twin simulators and cloud AI/ML platforms like AWS for testing the system's performance in various network scenarios. Validate the effectiveness of the multi-agent system in optimizing network parameters and handling complex tasks autonomously.
• Benchmarking and Analysis: Compare the autonomous system’s ability to adapt and optimize against traditional network management methods, focusing on metrics like response time, resource efficiency, and decision accuracy.

Skills Required

• A background in AI/ML with interest in reasoning retrieval-based models and multi-agent systems,
• Experience with RAG frameworks and large language models (LLMs),
• Proficiency in Python and AI libraries. Proficiency in backend development in general,
• Familiarity with containerization technologies like Docker and Kubernetes,
• Good understanding of cloud platforms, preferably AWS, and their AI/ML services,
• Fluent English language skills in both writing and conversation.(not strict).

Benefits for the Intern

This project offers a unique opportunity to work at the intersection of cutting-edge AI research and practical network management. The intern will develop skills in autonomous AI systems, cloud integration, and multi-agent orchestration, while being exposed to advanced concepts in AI reasoning and decision-making. Ericsson provides an environment rich with real-world challenges and the computational resources necessary to push the boundaries of AI applications in telecommunications.

Interns:

Interns who join us will enjoy an outstanding chance to make connections, to make change, to make a real difference. We are always looking for interns to become permanent members of our Ericsson team, so you’ll have the opportunity to discuss your future with us during your internship.

What happens once you apply?

Click here to find all you need to know about what our typical hiring process looks like. This may vary depending on the location and role.
Frequently asked questions
Students and Early Careers
Encouraging a diverse and inclusive organization is core to our values at Ericsson, that's why we nurture it in everything we do. We truly believe that by collaborating with people with different experiences we drive innovation, which is essential for our future growth. We encourage people from all backgrounds to apply and realize their full potential as part of our Ericsson team.
Ericsson is proud to be an Equal Opportunity and Affirmative Action employer, learn more.

Primary country and city: France (FR) || Massy

Req ID: 758091