Get Computer Vision Engineers
Vision ML engineers from Brazil, Argentina, and Colombia who build object detection, image classification, segmentation, and video analysis systems. They take models from training through to production inference — not just research notebooks.
Senior CV engineers in the US earn $185k-$240k. LATAM equivalents deliver at $82k-$125k — meaningful savings for roles that require both ML depth and systems engineering.
LATAM's 4-8 hour overlap with US teams supports collaborative annotation review and model evaluation sessions that benefit from synchronous feedback.
CV engineers bridge research papers and production systems. LATAM engineers in this space are used to communicating model accuracy vs inference latency tradeoffs to product teams.
LATAM universities produce computer vision researchers. Several engineers in our network have graduate CV backgrounds with publications or thesis work in object detection or medical imaging.
Vision applications often run on edge hardware. LATAM CV engineers increasingly have experience deploying optimized models to Jetson devices, Raspberry Pi, and mobile (Core ML, TFLite).
Every candidate completes all five stages before you see their profile. You can also run your own technical round after our screening.
A timed test measuring analytical thinking, pattern recognition, and problem-solving clarity, independent of specific programming language knowledge.
A structured interview assessing communication style, conflict resolution, ownership mindset, and English proficiency in a professional context.
A 90-minute live session covering three areas. First, model selection: given a detection task (objects, defects, or faces in a specific environment), candidates justify their model choice based on accuracy requirements, inference latency targets, and hardware constraints. Second, training setup: data augmentation strategy with Albumentations, handling class imbalance, mAP evaluation, and hyperparameter tuning approach. Third, production deployment: how they'd export a model to ONNX, run it through TensorRT, deploy to a Triton Inference Server, and set up a monitoring pipeline to catch performance degradation on a new input distribution.
Verification of work history, education, and identity with written consent, aligned with applicable privacy rules including LGPD where relevant.
We speak with at least two professional references who worked with the candidate in an engineering context, not personal contacts.
After our screening, you can optionally run your own technical round before making an offer.
Sample profiles
Anonymized profiles from our vetted talent pool. Actual candidates may vary.
São Paulo, Brazil
4+ years experience
Buenos Aires, Argentina
7+ years experience
Florianópolis, Brazil
10+ years experience
Want to see profiles matched to your specific requirements?
Request matched profilesWhat to expect
CV engineers in our network build systems that work in the real world, not just on benchmark datasets. They handle lighting variation, occlusion, domain shift, and latency requirements. They use AI tools to accelerate their own development workflow, particularly for annotation and architecture exploration.
Foundation models (SAM, Grounding DINO) are used to accelerate annotation workflows — auto-labeling reduces manual annotation time significantly on new datasets.
AI-assisted architecture search: they use tools like Copilot alongside papers-with-code to explore model variants faster before committing to a training run.
They define evaluation beyond mAP: they look at false positive rates by class, failure modes under specific lighting or camera conditions, and inference latency at the 99th percentile.
Model cards document training data distribution, failure modes, and intended deployment conditions. Future engineers don't discover edge cases in production.
They communicate production vs research performance gaps honestly. If a model performs 95% mAP on the test set but 78% on live video, they say so before deployment.
Junior CV Engineer
Mid CV Engineer
Senior CV Engineer
Lead CV Engineer
FAQ
Tell us your vision use case and deployment target. We'll have a shortlist in 3 days.
PyTorch
TensorFlow / Keras
ONNX export
Docker + Triton Inference Server