Lead Data Scientist

Role Mission

Our client is evolving their AI capabilities from traditional image processing to cutting-edge Deep Learning architectures. As a Lead Data Scientist, you will be the technical authority for our Computer Vision department. You will transition our existing "Imaging" workflows into high-performance, scalable Deep Learning models that solve high-impact problems for our clients.

Key Responsibilities

- Architectural Strategy:
Move beyond simple libraries to design custom neural network architectures (CNNs, Transformers, Diffusion models) tailored for specific business use cases.
- End-to-End Ownership:
Lead the full ML lifecycle—from data acquisition and annotation strategy to model deployment and monitoring.
- Technical Mentorship:
Act as the "North Star" for a team of 3-5 Data Scientists, conducting rigorous code reviews and fostering a culture of research-driven development.
- Cross-Functional Collaboration:
Partner with Product Managers and Stakeholders to translate vague business requirements into concrete technical roadmaps.
- Optimization & Scaling:
Ensure models are not just accurate, but efficient. You will oversee model compression (quantization, pruning) for cloud or edge deployment.

Required Technical Expertise

1. Deep Learning & Neural Architectures

- Deep expertise in
State-of-the-Art (SOTA)
models such as YOLOv8/10, Detectron2, Vision Transformers (ViT), and MAE (Masked Autoencoders).
- Solid understanding of the mathematical foundations:
- Optimization algorithms: $Adam, RMSProp, SGD$.
- Loss function customization: $Focal Loss, Dice Loss, Triplet Loss$.
- Experience with
Transfer Learning
and Fine-tuning Large Vision Models (LVMs).

2. Core Computer Vision (The Foundation)

- While we are moving toward DL, you must still be proficient in
OpenCV
for preprocessing: Filtering, Binary Morphology, Perspective/Affine Transformations, and Edge Detection.
- Experience in Feature Extraction and Image Segmentation (Semantic & Instance).

3. Software Engineering & MLOps

- Production Python:
Expert level; ability to write clean, modular, and testable code.
- Frameworks:
Mastery of
PyTorch
(preferred) or
TensorFlow/Keras
.
- Deployment:
Experience with Docker, Kubernetes, and model serving (FastAPI, NVIDIA Triton, or AWS SageMaker).
- Version Control:
Advanced Git workflows and CI/CD for Machine Learning (DVC, MLflow).

Qualifications & Experience

- Education:
BS or MS in Computer Science, Electrical Engineering, Mathematics, or a related field (or equivalent deep industry experience).
- Experience:
6-10 years in Data Science, with at least 3+ years specifically focused on Deep Learning and Computer Vision in a production environment.
- Leadership:
Proven experience leading a project from ideation to a revenue-generating or cost-saving production state.
- Communication:
Ability to explain complex $O(n)$ complexity or backpropagation logic to non-technical executive stakeholders.