Titulo Estágio
2025/26_N40 LightRAG in the Evolution of Telco OSS Products for Autonomous Operations
Local do Estágio
Altice Labs
Enquadramento
IDI - Áreas Chave
Artificial Intelligence/Gen AI
Tema / Titulo
LightRAG in the Evolution of Telco OSS Products for Autonomous Operations
In today's highly interconnected world, the telecommunications industry is undergoing a major transformation to meet the increasing demand forhigher bandwidth and more reliable connectivity. Traditional operations support systems (OSS) are evolving to incorporate advanced technologiesthat enable autonomous operations, reducing the need for manual interventions and increasing efficiency. As part of this evolution, the usage ofLLM / SLM is proving promising but, given the specificity of the telco business, its diverse technologies and the multiplicity of delivered services,the need to add some specific context is paramount. RAG may serve this purpose but have some limitations – namely Simplicity and LimitedCustomization, Ineffective Information Utilization and Static Knowledge Base -, that might be surpassed through LightRAG.
LightRAG offers several improvements over traditional RAG systems - Graph Structures, Dual-Level Retrieval, Incremental Updates and Efficiency –therefore promising to provide more accurate, contextually relevant and efficient responses.
This internship seeks to explore the potential of LightRAG in driving autonomous operations, proposing a set of possibilities and implementing aproof-of-concept scenario to demonstrate its effectiveness
Objetivo
The primary objectives of this internship are as follows:
Conduct a comprehensive literature review on current OSS products and the role of technologies like LightRAG in supporting autonomousoperations.
Study the specific challenges faced by telco OSS in terms of scalability, efficiency, and automation.
Develop a set of potential use cases where LightRAG can be integrated into OSS products to enhance autonomous operations.
Design customized models and algorithms using LightRAG to address identified challenges in OSS optimization and automation.
Implement a proof-of-concept scenario using real-world data to validate the proposed models and algorithms.
Evaluate the performance of the LightRAG-based solutions compared to traditional approaches, highlighting their advantages andlimitations.
Document the entire research process, findings, and recommendations for further improvements.
Provide practical solutions and insights that can be adopted by the industry to improve the efficiency and quality of telco services.
Plano de Trabalhos - Semestre 1
The project will be carried out in the following phases:
1: Literature Review and Domain Study
Conduct a thorough review of applicability of LightRAG in existing OSS products and technologies.
Identify key challenges and areas for improvement.
2: Proposal of Use Cases
Develop a set of potential use cases for integrating LightRAG into OSS products.
Analyze the feasibility and expected benefits of each use case.
3: Model Development
Design and develop customized models and algorithms using LightRAG.
Ensure that the models address the specific challenges identified in 1.
Plano de Trabalhos - Semestre 2
4: Proof-of-Concept Implementation
Implement the proposed models in a proof-of-concept scenario using real-world data.
Test and validate the performance of the models.
5: Analysis and Documentation
Analyze the results obtained from the proof-of-concept implementation.
Compare the performance of LightRAG-based solutions with traditional approaches.
Document the research process, findings, and recommendations.
6: Feedback and Refinement
Collect feedback from industry experts and stakeholders.
Make necessary adjustments and improvements to the models and algorithms.
7:Final Report
Condições
Integração no Programa GENIUS Investigação da Inova-Ria.
Entidade Promotora: Inova-Ria
Entidade de acolhimento: Altice Labs
Com possibilidade de integrar uma Bolsa de Investigação - Programa GENIUS - durante a realização do projeto de Dissertação - Integração numa equipa de I&D na Empresa Altice Labs .
Valor de bolsa de acordo com tabelas da FCT (ver pdf em (www.Inova-Ria.pt).
Período de realização de acordo com o enquadramento da Universidade.
• Horário: De acordo com enquadramento da Universidade
• Formato: (a combinar na entrevista)
• Meios: atribuição de um PC portátil e acessos à rede Interna da Altice Labs
• Kit de Acolhimento
• Onboarding nas equipas da Altice Labs, com atribuição de um tutor full-time
• Possibilidade de participar em todas as iniciativas de partilha de conhecimento ou de entretenimento levadas a cabo pela Altice Labs
• Possibilidade de entrada nos quadros da empresa
Os alunos interessados deverão enviar para genius@inova-ria.pt ao cc Dra Regina Maia Sacchetti (963618710).
• Curriculum Vitae;
• Disciplinas realizadas até ao momento com médias; simples documento eletrónico, que poderá obter no portal académico .
Processo de Seleção: Entrevista Inova-Ria - PROGRAMA GENIUS na qual fará parte o Orientador do projeto de forma a esclarecer a temática envolvida.
Observações
Aspetos Inovadores
Identification and validation of LightRAG benefits and its adequacy to OSS scenarios
Evaluation of applicability of LightRAG in a real OSS scenario
Evaluate LightRAG adequacy to support Autonomous Operations processes
Ferramentas a utilizar
Programming Languages: Python, Java
Data Processing: Pandas, SciKit Learn, PySpark
LLM ( large language models) and RAG (Retrieval-Augmented Generation)
Development Environment: Jupyter Notebooks, Linux Systems
Version Control: GIT
Project Management: JIRA, WIKI
Referências Bibliográficas
LightRAG in the Evolution of Telco OSS Products for Autonomous Operations
LightRAG, a lightweight and efficient RAG
LightRAG: Simple and Fast Retrieval-Augmented Generation
Vector RAG vs Graph RAG vs LightRAG
Orientador
Luís Cortesão
luis-m-cortesao@alticelabs.com 📩