Involve Space is an Italian aerospace company developing intelligent pseudo‑satellite platforms for Earth observation, defense, and scientific research. With our proprietary Stratostats® technology, we reach 30 km altitude to capture ultra‑high‑resolution data and deliver a new perspective on Earth.

We are now building CLEAR, our generative GeoAI platform, designed to turn this data into answers. CLEAR is meant to make territorial intelligence accessible through a conversational interface capable of understanding images, maps, and environmental phenomena, translating them into immediate insights and operational decisions.

The goal is simple yet ambitious: to create a mind that can “see, describe, and understand the planet” like never before.

In this context, the role of Generative AI Architect will be the technical and creative heart of CLEAR: the person who will design its cognitive foundations, defining the model architecture, training pipelines, and interactions among language, data, and vision.

They will also be responsible for the internal AI Adoption strategy, integrating models, agents, and automations into company workflows to build a truly AI‑native organization.

Responsibilities

Design and develop generative and multimodal AI architectures (text, images, geospatial data) supporting the CLEAR platform.
Build training and optimization pipelines for LLM, visual, and RAG models leveraging proprietary datasets derived from stratospheric missions.
Integrate reasoning, generation, and descriptive analysis capabilities on images and maps, transforming geospatial data into natural language.
Collaborate with data engineering and mission intelligence teams to create a scalable, high‑performance AI infrastructure, optimized for distributed and edge deployment.
Experiment with new generative AI architectures and methodologies, contributing to the company’s research and development roadmap.
Support the creation of demos and proofs of concept for partners, institutions, and stakeholders.
Lead AI adoption within the company, introducing intelligent agents and workflows into everyday processes.

Must Have

Proven experience (3+ years) designing advanced AI architectures, with a focus on generative and multimodal models.
Deep knowledge of frameworks such as PyTorch, TensorFlow, and Hugging Face Transformers.
Experience with fine‑tuning, prompt engineering, retrieval‑augmented generation (RAG), vector databases, and embedding models.
Solid foundation in machine learning, deep learning, and data processing pipelines (Python, APIs, Docker, cloud).
Knowledge of MLOps, distributed systems, and large‑scale AI architectures.
Experimental mindset, vision, and a pragmatic approach to innovation.
Excellent command of English (B2+).

Nice to have

Degree or academic background in Computer Science, Artificial Intelligence, Data Engineering, or related disciplines.
Experience in computer vision, multimodal reasoning, or geospatial AI.
Familiarity with GIS formats and tools (GeoTIFF, shapefiles, QGIS).
Ability to integrate and orchestrate cutting‑edge AI tools and models (LangChain, Llama Index, OpenAI, Anthropic, Gemini, Hugging Face, Perplexity, Mistral, Together AI, etc.).
Interest in aerospace, defense, environment, or data intelligence sectors.

What’s great in the job?

Joining a pioneering environment in aerospace and artificial intelligence.
Opportunity to work on frontier AI technologies integrated with real data from near space.
Direct collaborations with universities, research centers, and international industrial partners.
Professional growth opportunities in a rapidly expanding organization.
Permanent contract and salary aligned with experience.

What We Offer

Each employee has the opportunity to see the tangible impact of their work and contribute meaningfully to the company’s success. Throughout the year, various activities are often organized, such as events, team‑building moments, monthly aperitifs, and much more.