Mid–Senior Applied AI Research Engineer

We are teaching AI to understand "Space" (and it’s really hard!)

Role: Mid–Senior Applied AI Research Engineer

Type: Full-Time | Location: Tel Aviv (Hybrid)

Company: SWAPP.AI

Most "Generative AI" generates pixels or text. We generate Constructible Reality. 

SWAPP.AI is automating the architectural and engineering workflows end-to-end. We don't just make pretty pictures; our AI learns design languages and standards to generate precise, build-ready documentation, drawings, and models in record time.

The problem is massive: Architecture is a graph problem, a geometry problem, and a constraint-satisfaction problem. It is also a semantic challenge—where unstructured data (NLP), building codes, guidelines, and design intent must be translated into rigorous spatial rules.

We are looking for a Computer Scientist who specializes in AI, not just an ML researcher.

• You will thrive here if: You get excited by algorithms, Graph Neural Networks, Computational Geometry, Graph Theory, multimodal reasoning, and turning "spaghetti research" into robust production APIs.
• You will likely not enjoy this role if: You prefer to stay in Jupyter Notebooks and hand off code to "real engineers" to deploy.
• You will likely not enjoy this role if: You view code purely as a prototyping tool rather than a production asset.
  

You won't be just fine-tuning a model on clean data. You will be:

1. Reasoning with Geometry: Using Neural Networks and geometric deep learning to understand 2D/3D relationships and contexts.
2. Taming Multimodal Chaos: Ingesting messy drawings, text, and BIM models to create unified, consistent representations.
3. Shipping to Production: You own your stack—from the mathematical idea on the whiteboard to the CI/CD pipeline that deploys it.
   

We value strong CS fundamentals over knowledge of the latest flashy framework.

• Must Have: BSc/MSc in Computer Science (or equivalent). You know Big-O, data structures, and how systems behave in the real world.
• Must Have: 3+ years shipping ML to production (Industry experience).
• Must Have: Strong Python engineering and programming skills (testing, typing, modular design, version control).
• The Stack: PyTorch, GNNs, RL, Causal Inference, Vision/Geometry, and LLMs/NLP.
  

Send your CV to [email protected].

If relevant, include a link to your GitHub and/or a specific example of a "hard engineering problem" you solved (e.g., "I reduced inference time by 50% by rewriting the custom kernel" or "I built a pipeline to parse messy unstructured data into strict JSON").