Language models
Efficient training, supervised fine-tuning, alignment and evaluation.
Research
We publish at peer-reviewed venues and protect results with patents where it makes sense. What we learn feeds the product.
Efficient training, supervised fine-tuning, alignment and evaluation.
Hybrid architectures, semantic retrieval and grounding evaluation.
Detection, segmentation and edge deployment for industry.
Reproducible pipelines, observability and tests for AI systems.
Reproducible evaluation of LLMs in regulated domains
Itify Research, et al. · arXiv
Hybrid retrieval for industrial technical documentation
Itify Research, University Partner · ECAI
Efficient inline defect detection with compact models
Itify Research · Workshop @ NeurIPS
Contextualized retrieval system for AI applied to industry
Auditable evaluation method for generative models
We are open to industry challenges, EU proposals and academic collaborations.