The Mission
We are not looking for a generalist ML engineer to train standard models. We are looking for a deep-tech mathematical mind. At Swiftron, our foundation model relies purely on anonymized IDs with time deltas—no customer metadata, no product descriptions. Your mission is to extract maximum insight from these raw patterns and engineer highly complex geometric vector spaces (embeddings) that teach our neural networks exactly how a product behaves.
You will work hands-on on our core intellectual property, continuously improving the statistical foundation that feeds our predictive models.
Your Impact
- Vectorization & Combinatorics: Develop, combine, and optimize advanced statistical methods to transform sparse, anonymous transaction patterns into rich product embeddings.
- Push the Productive Boundary: While close to research, your ultimate goal is impact. Your algorithms will directly improve the accuracy of our live, production-grade AI model.
Generate New IP: Work explicitly on our patent-pending technology. You will be shaping the future of our Point Alignment Neural Network.
- Cross-Functional Execution: Leave the pure research bubble when needed. We expect you to jump into client projects alongside our Sales and Partner Managers to solve concrete data challenges.
Your DNA
- Mathematical Depth: You have deep, specialized expertise in high-dimensional vector spaces, combinatorics, logic, and advanced statistics.
- Tech Stack: You are highly proficient in Python, PyTorch, NumPy, and Jupyter Notebooks. You have a solid grasp of Keras/TensorFlow and know how to work efficiently with GPU clusters.
- Cloud Experience: Comfortable navigating and developing within Microsoft Azure and Google Cloud environments.
- The Greenfield Mentality: You love exploring uncharted territory. You are a builder who doesn't mind rolling up your sleeves for tasks slightly outside your job description when the team needs you.
What We Offer
- Expert Mentorship: Work side-by-side with our technical Co-Founder (former Head of AI and university lecturer) to push the boundaries of deep tech.
- PhD Collaboration: Pursuing an industrial PhD or academic teaching in a related field? We are highly flexible and open to offering this position as a 50% part-time role to support your academic goals.