AI / ML Engineer (Full Stack, Platform & Infrastructure)

SPACE N PLACE is an international AI startup in the PropTech sector. We create two interconnected systems: * An AI platform for real estate analysis (Computer Vision, LLM, numerical metrics, visual layers) * A smart real estate search engine that matches properties to user life scenarios and analyzes them before clicks. Our goal is to turn the chaotic market of listings into a structured, analyzable system and help people make informed decisions. We build the product as a scalable AI platform, not an MVP 'for speed'.

### Responsibilities - Develop and maintain the AI/ML part of the product (LLM, RAG, Computer Vision) - Design and implement data pipelines and data ingestion - Work with vector databases and semantic search - Integrate and orchestrate various AI models (LLM, CV, multimodal) - Deploy and maintain models on servers - Optimize costs for model operation and computational infrastructure - Design and build computing clusters for scaling load - Develop backend infrastructure (FastAPI, async, high-load scenarios) - Optimize performance and latency of AI pipelines - Work with queues, caching, and real-time data processing - Support and develop existing architecture (Docker, API, services) - Participate in system design at the platform level, not just individual tasks We are looking for an engineer who understands how to build, deploy, and maintain an AI system in production, not just integrate APIs.

### Requirements - Practical experience with AI/ML systems in production - Experience building RAG, LLM integrations, or data pipelines - Strong backend skills in Python (FastAPI, async) - Understanding of high-load system architecture - Experience with vector databases or semantic search - Experience deploying and maintaining models on servers - Ability to optimize latency, cost of AI services, and computations - Understanding of how to build and scale computing clusters under load - Independence and systematic thinking

### Conditions - Remote work - Part-time / project format - Hourly payment (discussed individually) - Tasks and volume agreed upon in advance with time estimation - Focus on results, not just occupancy - Work on real AI tasks (not just 'wrapping APIs') - Direct influence on architecture, models, and product