Propostas Submetidas

DEI - FCTUC
Gerado a 2025-07-17 14:48:31 (Europe/Lisbon).
Voltar

Titulo Estágio

Lightweight AI Models for In-House Deployment

Áreas de especialidade

Sistemas Inteligentes

Local do Estágio

Coimbra

Enquadramento

Not all AI tasks require Large Language Models (LLMs) — especially when latency, cost or privacy are concerns. This internship focuses on exploring and prototyping open-source ML and Small Language Model (SLM) solutions that can be hosted internally. The goal is to replace high-frequency third-party LLM calls with efficient in-house models, offering better control, lower cost and multilingual support. These models will support real-time, cost-sensitive, and privacy-critical applications.

Objetivo

- Benchmark open-source models in English and at least one additional language
- Prototype and document deployment pipelines using tools like TorchScript, ONNX, and Triton
- Evaluate real-time performance and privacy benefits across use cases
- Design a scalable architecture for multilingual inference services
- Propose hybrid integration strategies for AI feature pipelines
- Reduce inference costs while maintaining or improving response time and accuracy

Plano de Trabalhos - Semestre 1

- Survey and select candidate models for sentiment, NER, summarization, etc.
- Benchmark selected models in English and one additional language (e.g., Spanish or French)
- Prototype sample deployments using TorchScript, ONNX, or Triton
- Implement language detection and pipeline branching mechanisms
- Document performance and cost tradeoffs for internal hosting vs third-party LLMs

Plano de Trabalhos - Semestre 2

- Integrate models into example applications (e.g., redaction, post-call summaries, intent routing)
- Develop a scalable architecture for multilingual inference services
- Evaluate hybrid strategies combining local models with LLM-based fallbacks
- Propose integration guidelines for AI feature teams across product lines
- Deliver final presentation and deployment documentation

Condições

- Monthly stipend of €900, based on a full-time commitment (40 hours per week)
- Hybrid work model: interns are expected to work from the company office 2–3 days per week
- PC or Mac for the duration of the internship
- Integration into a Product Area team, working closely with engineers and stakeholders
- A mentor will be assigned to support the intern’s development and success

Observações

What We Are Looking For:
- Self-motivated individuals with a strong desire to learn and a collaborative mindset
- Final-year Master’s students in Computer Science, Artificial Intelligence, Engineering or related fields
- A solid understanding of software architecture fundamentals
- Demonstrated experience with at least one relevant programming language or technology (e.g., Python, Java, .NET or C/C++)
- A genuine interest in applied machine learning and real-time AI systems
- Familiarity with NLP models, deployment tools (e.g., ONNX, TorchScript), or multilingual AI is a plus

Orientador

Mariana Dias
mdias@intermedia.com 📩