Principal ML & AI Engineer (Spatial AI & Perception)
HERE Technologies22 hours ago
Berlin, Germany +3 moreStaff+
Responsibilities
- Develop and train infrastructure perception models for lane detection, road boundaries, traffic signs and lights, and road-surface attributes.
- Build BEV-based scene-representation and occupancy-network models, integrate geospatial priors, and validate synthetic training data.
- Run large-scale distributed training and implement scalable data-loading, augmentation, experimentation, and validation pipelines.
- Own model optimization for edge deployment, including export, compilation, operator compatibility, graph optimization, quantization, compression, and accuracy recovery.
- Profile and benchmark models on embedded hardware for latency, throughput, memory use, and bottlenecks.
- Translate state-of-the-art research into reproducible training pipelines and conduct experiments, A/B tests, and ablation studies.
- Partner with research, simulation, and technical leadership on architecture, quantization, precision, and deployment decisions.
- Maintain code quality, documentation, extensibility, and knowledge sharing while mentoring junior engineers.
Requirements
- 5–7 years of experience in ML/AI, including at least 3 years in computer vision, perception, or deep learning systems.
- Experience shipping production perception models into real-world systems or large-scale data pipelines.
- Hands-on experience deploying and optimizing models on embedded hardware or automotive-grade SoCs.
- Strong PyTorch expertise, including custom architectures, loss functions, training loops, DDP/FSDP, mixed precision, and gradient checkpointing.
- Experience with large-scale distributed training beyond fine-tuning pretrained models.
- Experience with BEV or multi-camera transformer architectures such as BEVFormer, BEVDet, BEVFusion, or occupancy networks.
- Experience with infrastructure perception, structured scene understanding, semantic or instance segmentation, and BEV-space, occupancy-grid, or map-aligned representations.
- Hands-on expertise with ONNX and TensorRT/QNN, quantization workflows such as QAT and PTQ, mixed-precision inference, per-layer sensitivity analysis, and model compression.
- Familiarity with Qualcomm Snapdragon Ride, NVIDIA Orin, TI TDA4, or comparable embedded accelerators, including real-hardware profiling and inference-throughput optimization.
- Ability to evaluate coverage, failure modes, edge cases, regression behavior, accuracy-latency trade-offs, and deployment readiness.
- Ability to reproduce research papers, adapt architectures, and integrate them into scalable production-ready training loops.
Benefits
- Hybrid work arrangement with 2 days in the office and 3 days from home each week.
- 30 paid vacation days, flexible working hours, and work-life-balance support.
- On-site gym and sauna, yoga room, wellness programs, and life-coaching sessions for applicable Berlin employees.
- Opportunity to work on high-scale services serving and storing petabytes of data.
- Competitive salary plus bonus.
- BVG ticket and German language course for applicable Germany-based employees.
- Employee events, brown-bag talks, rooftop BBQs, and a diverse international team.
Tech Stack
MLflowPyTorch
Categories
About HERE Technologies
HERE Technologies builds a location data and mapping platform used by automakers, logistics providers, and enterprises to power navigation, routing, and geospatial analytics. It sells map data, APIs/SDKs, and platform services for fleet, mobility, and supply-chain applications. The company traces its roots to Navteq (founded 1985) and Nokia Maps, and is privately held with headquarters in Eindhoven, Netherlands.