← Back to Jobs
AI and Machine Learning
Bengaluru
3-5 Years
Full-time
On-site
Apply Now →
Deep Learning Engineer — AI -Computer Vision
AI and Machine Learning
Bengaluru
3-5 Years
Full-time
On-site
Client Project Opportunity
We want a technically strong deep learning engineer who has put computer vision models / VLMs into real production, on real camera streams. You will own video analytics end to end — building and training models, optimizing them for inference, and running them live against multi-camera deployments. Hands-on role, heavy Python and C++, live systems.
About the Role
We want a technically strong deep learning engineer who has put computer vision models / VLMs into real production, on real camera streams. You will own video analytics end to end — building and training models, optimizing them for inference, and running them live against multi-camera deployments. Hands-on role, heavy Python and C++, live systems.
Key Responsibilities
- Build, train and evaluate computer vision models for detection, tracking, classification and segmentation.
- Design and deliver end-to-end video analytics pipelines running on live camera streams.
- Build and optimize multi-camera tracking across overlapping and non-overlapping views.
- Apply Vision Language Models to real-world scene-understanding problems, including prompting, grounding, evaluation and pipeline integration.
- Optimize models for real-time inference on production GPU and CPU server hardware.
- Build and automate data-processing pipelines covering ingestion, annotation handling, augmentation and dataset versioning.
- Debug and improve model performance against challenging real-world footage, including poor lighting, occlusion, camera drift and varied camera angles.
- Deploy, monitor and retrain models in production.
Must-Have Skills & Experience
- 3–5 years of experience building and deploying deep learning models in production.
- Strong programming skills in Python and C++.
- Hands-on experience with PyTorch and TensorFlow.
- Experience with DeepStream and OpenVINO.
- Proven experience building production video analytics pipelines.
- Experience with multi-camera tracking, including re-identification, cross-camera association and identity handover.
- Strong computer vision knowledge covering CNNs, the YOLO family and Vision Transformers.
- Practical experience using Vision Language Models on real-world problems.
- Deep understanding of image-processing pipelines, including colour spaces, codecs, decoding, resizing, normalization and batching.
- Experience working with real-world camera streams using RTSP, ONVIF, different resolutions and multiple camera vendors.
- Experience processing live video streams at production scale.
- Experience deploying models on production-grade NVIDIA data-centre GPUs and Intel Xeon servers.
- Strong data-manipulation, processing and pipeline-automation skills.
- Strong understanding of deep-learning fundamentals and model architecture behavior.
- Experience with Docker, Git and DVC.
Good to Have
- Experience with Large Language Models and Retrieval-Augmented Generation pipelines.
- Experience with multi-agent workflows.
- Experience with workflow-orchestration platforms.
- Experience with Kubernetes.
- Experience with TensorRT and ONNX.
- Knowledge of NIST Face Recognition Vendor Test standards.
Role Details
- Experience - 3-5 Years
- Location - Bengaluru, India
- Work Mode - On-site
- Employment - Full-time
- Deployment - Client Project
- Domain - AI / Computer Vision / Video Analytics
Apply for this Job
* indicates a required field