Senior ML Engineer

Job Description

General Info

We are developing a breakthrough imaging and 3D technology platform currently under provisional patent. Our Budapest engineering team works at the edge of what mobile hardware and modern AI make possible, in a well-funded, non-corporate environment built around craftsmanship and technical excellence.

As Senior Machine Learning Engineer, you will be a deeply technical individual contributor responsible for designing, training, evaluating, and productionising the ML and deep-learning models at the core of our platform — including computer vision, 3D reconstruction, and related imaging capabilities. You will work closely with the Lead Machine Learning Engineer on ML strategy and execution, partner with the Lead Infrastructure Engineer on the training-to-serving handoff, and collaborate with the Lead Mobile Software Engineer on client-side ML capabilities. This role is for an experienced engineer who can take significant areas of the model stack and own them end to end, with a high bar for scientific rigour and engineering craft.

Key Responsibilities

Model Development & Research

• Design, train, and ship advanced ML and deep-learning models powering the Company’s core technology, including computer vision, 3D reconstruction, and related imaging capabilities.
• Take ownership of significant model components — architecture choices, loss functions, and training recipes — under the technical direction set by the Lead ML Engineer.
• Read, critique, and replicate relevant papers; prototype promising approaches; and apply external research selectively and rigorously to internal problems.
• Contribute to the Company’s evaluation methodology: metric design, test sets, held-out benchmarks, failure-mode analysis, and uncertainty quantification for imaging models.

Data, Training & Productionisation

• Execute the end-to-end training workflow for owned models: dataset curation and labelling, large-scale training on the Company’s on-premises GPU cluster, distributed training, mixed-precision, and hyperparameter search.
• Drive inference optimisation for owned models — quantisation, pruning, distillation — and hardware-specific acceleration, with clean handoffs to the Infrastructure team for cloud serving.
• Apply experiment tracking, reproducibility, and versioning practices rigorously, so that every production model can be traced back to its training provenance.
• Help define monitoring, drift detection, and rollback processes for production models, working with the Infrastructure team on the observability and incident-response substrate.

Cross-Functional Collaboration

• Partner with the Lead Infrastructure Engineer on model export, versioning, staged rollouts, A/B testing, and GPU-efficient serving patterns.
• Partner with the Lead Mobile Software Engineer on on-device or hybrid ML capabilities, including model compression for mobile deployment and the cloud-to-mobile inference interface.
• Work with product and design to translate model capabilities, constraints, and failure modes into concrete product decisions.
• Contribute to IP-related discussions where novel methods may warrant patent filings.

Engineering Craft

• Run and participate in technical design reviews, paper-reading sessions, and model evaluations.
• Maintain high-quality, living documentation for owned models — architectures, training procedures, evaluation results, and known failure modes.
• Mentor more junior engineers on owned areas as the team grows.
• Hold a high bar for scientific honesty about what models actually do and do not do, in production conditions.

Qualifications

Required

• Master’s degree in Computer Science, Machine Learning, Electrical Engineering, Imaging Science, Physics, or a closely related technical field.
• Five to seven years of ML engineering experience, including hands-on work that has taken models from research into production.
• Demonstrated track record of shipping production ML systems — with the operational experience that comes from running models through training, evaluation, deployment, monitoring, and iteration cycles.
• Strong proficiency in Python and modern ML frameworks (PyTorch preferred; TensorFlow, JAX, or similar welcomed).
• Solid hands-on experience training models on GPU clusters, including the NVIDIA stack (CUDA, cuDNN, NCCL), distributed training (data and model parallelism), and mixed-precision training.
• Working knowledge of inference optimisation — quantisation, pruning, distillation — and hardware-aware acceleration (ONNX, TensorRT, CoreML, or similar).
• Strong research instincts: ability to read, critique, and replicate papers; to distinguish signal from hype; and to apply external research selectively and rigorously to internal problems.
• Ability to handle confidential and pre-patent technical material with discretion, and to follow the Company’s IP and data-handling policies rigorously.
• Fluency in English (written and spoken); able to work full-time in Budapest, Hungary.

Strongly Preferred

• PhD in a relevant research area (computer vision, 3D reconstruction, deep learning, or related).
• Peer-reviewed publications, substantive open-source contributions, or equivalent externally visible research output.
• Direct experience with computer vision, 3D reconstruction, neural rendering, or related imaging ML domains — including shipped products or substantive research contributions.
• Experience with real-time or near-real-time inference, mobile-to-server ML workflows, or sensor-based modelling.
• Experience in early-stage startup or innovation-lab environments; track record of contributing to first-of-its-kind products or platforms.

Compensation

In accordance with Hungary’s implementation of the EU Pay Transparency Directive (Directive (EU) 2023/970):

• Salary range: HUF 2,000,000 – HUF 2,500,000 gross per month, depending on experience and demonstrated technical depth.
• Discretionary performance bonus, with details discussed with shortlisted candidates.
• Statutory and customary Hungarian benefits.

Compensation decisions are made on the basis of objective, gender-neutral criteria, including relevant experience, technical skills, scope of responsibility, and demonstrated performance. The Company does not request salary history during the recruitment process.