Job Description:

This role blends hands-on data labelling, model experimentation, and performance tuning to ensure the platform delivers robust and real-time vision analytics.

We are looking for an AI Model Developer who will build generalized computer vision models for camera imagery. You will play a key role in developing, fine-tuning, and optimizing models that can generalize across diverse camera feeds and environments. The role requires strong ownership of the full AI lifecycle — from data annotation planning to model deployment — and a passion for experimenting with different strategies to push model performance boundaries.

Key Responsiblities

- Model Development: Develop and optimize deep learning models for camera images, focusing on generalization across varied domains.
- Hyperparameter Tuning: Conduct systematic experiments with model architectures and hyperparameters to achieve the best trade-off between accuracy, robustness, and speed.
- Annotation Management:

- Plan weekly annotation activities and strategies for the annotation team.
- Review and approve annotations to ensure quality and consistency.
- Design new labeling strategies to improve model understanding of complex scenarios.

Model Training & Evaluation:

- Train and evaluate models on curated datasets.
- Compare new model performance against previous baselines to identify improvement trends.
- Analyze model failure cases and design corrective actions.

Automation & MLOps:

- Develop and maintain data/model pipelines to automate repetitive or manual workflows.
- Implement experiment tracking, versioning,



and automated training/evaluation loops.

Experimentation & Research:

- Explore new architectures, augmentation techniques, and annotation strategies.
- Stay updated on recent trends in multimodal AI and apply them to complex surveillance use cases.

Requirements :

- Bachelor’s or Master’s degree in Computer Science, AI/ML, or related field.
- Strong experience with PyTorch or TensorFlow for model development.
- Hands-on experience in computer vision tasks such as detection, segmentation, or classification.
- Proficiency in hyperparameter tuning and training large-scale models.
- Familiarity with data annotation tools like Roboflow or CVAT and managing annotation quality.
- Experience in building ML pipelines and MLOps workflows (e.g., MLflow, DVC, Kubeflow).
- Analytical mindset for model evaluation and performance benchmarking.
- Excellent teamwork and communication skills.

Good-to-Haves

- Experience with multi-camera or multimodal data (e.g., image + metadata).
- Familiarity with Edge AI or real-time video inference.
- Knowledge of data versioning, model drift detection, or continuous learning.
- Prior experience managing or coordinating a small annotation team.

What we Offer

- 20 GB GPU server access for experimentation and training.
- Lead and coordinate your own annotation team.
- Work directly with the Co-founder and Senior Developer on high-impact projects.
- Exposure to multimodal, real-world AI problems that go beyond typical datasets.
- Fast-paced workplace with room for innovation, autonomy, and career growth.

📌 AI Model Developer - Computer Vision (Rajkot)
🏢 Mantras2Success
📍 Rajkot