Occasional remote work possibility

Job Description

We’re looking for a Senior AI/Computer Vision Engineer to build AI systems end to end, from defining the problem and data requirements to training models, validating them, deploying on edge devices, and integrating them into C++ pipelines.

This is a hands-on engineering role for someone who can take applied AI projects from concept to production and is comfortable working across modeling, evaluation, deployment, and system integration.

Own AI and computer vision projects end to end, from technical definition to production deployment;
Translate product needs into system, data, model, and evaluation requirements;
Define data collection, annotation, and validation strategies;
Train and optimize deep learning models for real-world vision and multimodal applications;
Build testing, benchmarking, and failure analysis workflows;
Optimize models for edge devices with attention to latency, memory, throughput, and accuracy tradeoffs;
Integrate models into production inference systems, including C++ edge pipelines;
Work closely with software and product teams to ship reliable AI features at scale.

Requirements

MS or PhD in Computer Science, Electrical Engineering, or a related field, or equivalent practical experience
Strong Python skills
Solid experience with PyTorch, TensorFlow, or similar frameworks
4+ years of experience building and shipping computer vision and/or applied AI systems
Hands-on experience with tasks such as detection, classification, segmentation, tracking, embeddings, or multimodal perception
Experience beyond model training, including evaluation, deployment, and production support
Experience optimizing models for edge or embedded environments
Strong software engineering skills, including C++ production pipelines or inference systems
Fluent in English and Latvian

Nice to have

Experience with LLMs or multimodal models in production
Experience with ONNX, TensorRT, OpenVINO, or TFLite
Experience with MLOps and experiment tools such as MLflow, Weights & Biases, or DVC
Familiarity with model monitoring and post-deployment analysis