Pere Martra
AI Engineer & Applied Researcher specializing in LLM optimization, structured pruning, and fairness-aware systems.
I help ML engineering teams optimize large language models for production through practical research and hands-on implementation.
Current Work
- Writing Rearchitecting LLMs for Manning Publications
- Developing frameworks for agent migration to open-source models
- Active Open Source contributor (1,700+ GitHub stars)
- Developing OptipFair, a fairness-aware pruning library for LLMs.
Training
Hands-on workshops for engineering teams on LLM optimization and creation.
What I Offer
- LLM optimization techniques (pruning, quantization, distillation)
- Fairness-aware model development
- Production deployment strategies
Format
Short workshops • Small teams • Customized content
Limited availability: 1-2 engagements maximum
Consulting
Specialized consulting services for production LLM challenges.
LLM Optimization
Hands-on optimization for production deployments. Structured pruning, quantization, and efficiency improvements.
Agent Migration
Research Partnership
Collaborative framework development for migrating agentic systems from proprietary models (GPT-4, Claude) to open-source alternatives (Llama, Mistral).
Limited availability: 1-2 engagements maximum
Selected Work
Books
Rearchitecting LLMs (Manning, forthcoming)
Large Language Models Projects (Apress, 2024)
Open Source
OptiPfair – Fairness-aware structured pruning
LLM course – LLMs: Hands-on Projects
1,700+ GitHub stars combined