Foundational AI Engineer — Build In-House Models (Remote)

Maisa; Madrid, Barcelona, Valencia, Murcia; Hybrid; GBP 90,000 - 130,000; Full time

Job summary

Maisa is seeking a Foundational AI Engineer to join our AI team, reporting to the co‑founder and Chief Science Officer. You will train, fine‑tune, post‑train, and evaluate our own language models, with a focus on reliable production use in regulated environments. You’ll work hands‑on in Python and PyTorch toward scalable in‑house solutions.

This early‑stage startup values ownership and direct impact. You’ll collaborate with Applied AI, product, and engineering teams, shaping our model strategy

Qualifications

• Hands-on experience training and fine‑tuning models.
• Rigorous evaluation and ability to explain model quality.
• Strong Python programming and software engineering fundamentals.
• Ability to operate with ownership in a fast-paced environment.
• Clear communication with engineering, product and leadership.

Responsibilities

• Train, fine‑tune, and post‑train Maisa's own language models from data prep to release.
• Design and run rigorous model evaluation reflecting real client use cases.
• Build and improve training pipelines, data curation, tracking, and reproducibility.
• Apply modern training techniques and assess their real impact in production.
• Collaborate with AI infrastructure on model serving and inference optimization inside client environments.
• Monitor models for quality drift and trigger retraining as needed.
• Contribute to shaping our model strategy and align with product needs.

Skills

Hands-on model training

Evaluation methodology

Python software engineering

Communication with engineers/product

Ownership in fast-moving environment

Tools

PyTorch

Python

Model serving (vLLM/SGLang)

Hugging Face ecosystem

Distributed training

Job description

Maisa is seeking a Foundational AI Engineer to join our AI team, reporting to the co‑founder and Chief Science Officer. You will train, fine‑tune, post‑train, and evaluate our own language models, with a focus on reliable production use in regulated environments. You’ll work hands‑on in Python and PyTorch toward scalable in‑house solutions.

This early‑stage startup values ownership and direct impact. You’ll collaborate with Applied AI, product, and engineering teams, shaping our model strategy