This is us

At Avenga, we believe that human creativity empowers technology that matters. Operating globally, our 6000+ specialists provide a full spectrum of services, including business and tech advisory, enterprise solutions, CX, UX and Ul design, managed services, product development, and software development.

This is the job

In Bulgaria we are actively seeking a Senior Embedded Engineer (IoT) to strengthen our team.

This is you

• Python (production-grade): NumPy, Pandas, PyTest, type hints; solid debugging and profiling.
• ML fundamentals: supervised learning, metrics (accuracy, F1, mAP, AUROC), regularization, cross-validation, calibration.
• Deep Learning: PyTorch (preferred) or TensorFlow/Keras; training loops, checkpoints, mixed precision, schedulers.
• 2D CV basics: CNNs/ViTs, data augmentation, OpenCV; classification & detection (e.g., ResNet/EfficientNet, YOLO/Detectron2).
• 3D basics: point cloud fundamentals (voxelization, downsampling, ICP), simple 3D classifiers/detectors (PointNet/PointNet++, OpenPCDet or MMDetection3D exposure).
• MLOps hygiene: Git, experiment tracking (TensorBoard/Weights & Biases/MLflow), model versioning, reproducible environments.
• Production mindset: logging/monitoring, latency/throughput tradeoffs, error budgets, clear documentation.
• Industrial context: understanding of shop-floor data (PLC tags, historian trends), basic safety & uptime awareness.

Nice-to-have skills:

• 3D perception: MinkowskiEngine, sparse CNNs, voxel/BEV pipelines, NMS tuning, class imbalance strategies.
• Edge deployment: NVIDIA Jetson/TensorRT, ONNX conversion, quantization (INT8), CUDA fundamentals.
• Time-series & anomaly: forecasting, change-point detection, isolation forests/autoencoders.
• Agents/LLM-adjacent: RAG for tech docs/SOPs, tool-calling, simple operational assistants for operators/maintenance.
• Pipelines: Kafka/MQTT, Airflow/Prefect, Streamz/Faust; REST/gRPC services.
• IIoT & platforms: Ignition scripting (Python/Jython), ThingWorx services, Kepware tag models, OPC UA/MQTT structures.
• Datastores: TSDB (InfluxDB/TimescaleDB), PostgreSQL, object storage; basic schema design.
• Containers/CI: Docker, Compose, basic GitHub Actions/GitLab CI.
• ROS/ROS2 exposure for AGV/robotic data; calibration and sensor fusion basics.
• Quality systems: familiarity with PPAP, SPC, traceability; (bonus) safety standards awareness (IEC 61508/ISO 26262).

This is your role

• Design, train, and evaluate 2D (images/video) and 3D (point clouds/LiDAR) models for defect detection, object classification, pose/volume estimation, and anomaly detection.
• Prepare datasets (ingest, labeling strategy, augmentation, balancing, train/val/test splits) and own experiment tracking & reproducibility.
• Build inference pipelines from sensors to edge/cloud (pre-processing, model serving, monitoring, drift alerts).
• Integrate models with Ignition/ThingWorx and industrial data buses (OPC UA, MQTT, REST) including real-time triggers and historian/TSDB writes.
• Collaborate with manufacturing engineers/OT teams; translate process specs and acceptance criteria into measurable ML KPIs.
• (Senior) Own architecture & roadmap; mentor juniors; drive stakeholder alignment and production rollouts.

What awaits you at Avenga?

Through our values, Better Minds, Bolder Ideas and Bigger Hearts, we strive to provide you with the tools, the autonomy, the trust, and assistance you need to excel. Enjoy benefits like private health insurance, well-being programs, flexible and hybrid work models, laptops and gear, trainings, language classes, social events, great offices, and more.

We take pride in the diverse skills and character of our teams, welcoming everyone to apply and contribute to our collective strength.