Principal Computer Vision Scientist

SOURCEABILITY SG PTE. LTD.; Singapore; On-site; SGD 180,000 - 300,000; Full time

Job summary

Sourceability seeks a Principal Computer Vision Scientist to lead advanced CV and AI/ML work within GEO. This senior role guides research direction, model architecture, experimentation, and production readiness for CV solutions used in company products.

The role requires a PhD and hands-on CV/ML experience, comfortable with production systems, imperfect data, and evolving product requirements. You will mentor engineers and collaborate across product, engineering, and QA to deliver

Qualifications

• PhD in Computer Science, Computer Vision, ML, AI, math, or related field.
• 7+ years hands-on ML/DL with strong CV focus.
• Strong practical experience with PyTorch and/or TensorFlow.
• Strong Python development skills.
• Experience with OpenCV, NumPy, Pandas, scikit-learn and modern Python ML ecosystem.
• Deep understanding of CV algorithms and DL approaches.
• Experience with object detection, segmentation, classification, feature matching, retrieval, evaluation.
• Experience with CNNs, Transformers, Vision Transformers, YOLO, Mask R-CNN, CLIP-like and SAM-like models.
• Experience bringing ML models into production and optimizing for inference.

Responsibilities

• Lead research, design, development, and implementation of CV/AI/ML solutions.
• Define model architecture, approach, experiment strategy, validation, and production readiness.
• Train, fine-tune, evaluate, optimize, and deploy CV/ML models.
• Own full model lifecycle: data, datasets, annotation, training, evaluation, deployment, monitoring.
• Build prototypes and demos to validate new ideas before full product implementation.
• Review and improve CV pipelines, model quality, inference performance, and reliability.
• Collaborate with engineers to integrate models into production apps/services.
• Define standards for evaluation, versioning, data management, reproducibility, and MLOps.

Skills

PyTorch

Python

OpenCV

TensorFlow

NumPy

Pandas

scikit-learn

CNNs

Transformers

MLOps

REST APIs

Docker

CI/CD

Cloud ML

Distributed teams

Agile

Model evaluation

Education

PhD in Computer Science / CV / ML

Postgraduate research in CV

Tools

PyTorch

TensorFlow

OpenCV

NumPy

Job description

Sourceability is building a new Global Engineering Organization (GEO) to strengthen internal software delivery, improve production ownership, and build long-term engineering capability inside the company.

We are looking for a Principal Computer Vision Scientist to lead advanced Computer Vision and AI / ML work inside GEO. This role will be responsible for research direction, model architecture, experimentation, model quality, production readiness, and practical implementation of computer vision solutions used in company products.

This is a senior technical leadership role for a highly experienced specialist who can work across research, engineering, product, and production systems. The right candidate should be able to evaluate new approaches, design model architectures, run experiments, improve model quality, and help engineering teams bring AI / ML capabilities into real production workflows.

This role requires PhD-level education and strong hands‑on experience in applied Computer Vision, Machine Learning, and Deep Learning. The person in this role should be comfortable working with business‑critical systems, practical production constraints, imperfect datasets, and evolving product requirements.

Assigned Product Group

This role will be primarily aligned with the Computer Vision product group inside GEO.

The role may also support other internal product groups or AI / ML initiatives where computer vision, image processing, visual search, object detection, segmentation, classification, or model evaluation expertise is needed.

Product Group Focus Areas

Depending on business priorities, the role may focus on one or more of the following areas:

• Mobile App / Computer Vision: Image capture workflows, mobile application integration, computer vision model development, model inference, warehouse / field usability, user feedback loops, production model quality, and UAT support.
• AI / ML Product Features: Applied machine learning features, model evaluation workflows, proof‑of‑concepts, model‑assisted automation, AI‑assisted business tools, and integration of AI / ML capabilities into existing business workflows.
• Data and Annotation Workflows: Image datasets, data quality, annotation requirements, labeling guidelines, model training datasets, validation datasets, failure case analysis, and continuous improvement of model performance.
• Production ML Systems: Model deployment, model versioning, inference performance, monitoring, reproducibility, scalability, reliability, and MLOps practices.

Insight on Your Impact

In this role, you will:

• Lead research, design, development, and implementation of Computer Vision and AI / ML solutions.
• Define model architecture, technical approach, experiment strategy, validation methodology, and production readiness criteria.
• Train, fine‑tune, evaluate, optimize, and deploy models for object detection, semantic segmentation, image classification, feature matching, OCR, visual search, and image understanding.
• Own the full model lifecycle, including data analysis, dataset quality, annotation requirements, model training, experiment tracking, evaluation, deployment, monitoring, and continuous improvement.
• Build prototypes, proof‑of‑concepts, demos, and technical experiments to validate new ideas before full product implementation.
• Analyze model performance, identify failure cases, and recommend practical improvements based on data, user behavior, and business needs.
• Review and improve existing Computer Vision pipelines, model quality, inference performance, scalability, and production reliability.
• Work with software engineers to integrate ML models into production applications and services.
• Define standards for model evaluation, model versioning, dataset management, reproducibility, and MLOps practices.
• Evaluate research papers, open‑source models, AI platforms, and new technologies for potential use in company products.
• Provide technical guidance and mentoring to engineers working on AI / ML and computer vision features.
• Support planning and estimation for AI / ML work by clarifying technical complexity, risks, dependencies, and realistic delivery assumptions.
• Create technical documentation, model evaluation reports, architecture notes, and recommendations for engineering and product teams.
• Partner with Product / Delivery Managers to translate business needs into practical AI / ML implementation plans.
• Partner with Engineering Managers, Team Leads / Architects, QA, DevOps, Data, and business stakeholders to make sure AI / ML work can be delivered and supported in production.

Your Qualifications, Your Influence

To be successful in this role, you should have:

• PhD in Computer Science, Computer Vision, Machine Learning, Artificial Intelligence, Applied Mathematics, Electrical Engineering, Robotics, or closely related technical field.
• 7+ years of hands‑on experience in Machine Learning / Deep Learning, with strong focus on Computer Vision.
• Strong practical experience with PyTorch and / or TensorFlow.
• Strong Python development skills.
• Experience with OpenCV, NumPy, Pandas, scikit‑learn, and modern Python ML ecosystem.
• Deep understanding of classical Computer Vision algorithms and modern deep learning approaches.
• Strong experience with object detection, semantic segmentation, image classification, feature matching, image retrieval, and model evaluation.
• Experience with modern Computer Vision architectures and techniques, including CNNs, Transformers, Vision Transformers, YOLO, Mask R-CNN, CLIP‑like models, SAM‑like models, or similar.
• Experience bringing ML models into production environments.
• Experience with model optimization for inference speed, latency, memory usage, scalability, and reliability.
• Experience with REST APIs, Docker, CI / CD, model versioning, experiment tracking, and MLOps practices.
• Strong understanding of datasets, data quality, annotation processes, labeling requirements, and model error analysis.
• Ability to read, understand, and evaluate technical documentation and research papers in English.
• Ability to explain complex technical topics to engineering, product, and business stakeholders.
• Experience working in Agile software development environment.
• Strong ownership mindset, good judgment, and ability to make practical technical decisions under uncertainty.
• Comfortable working in distributed teams across multiple locations and time zones.

Preferred Skills and Technical Familiarity

The following experience will be helpful:

• Post‑PhD research or industry experience in applied Computer Vision.
• Publications, patents, or strong applied research record in Computer Vision, Machine Learning, or AI.
• Experience leading technical direction for AI / ML projects.
• Experience mentoring ML engineers, software engineers, or data annotation teams.
• Experience with edge or mobile inference technologies, including ONNX, TensorRT, OpenVINO, TFLite, CoreML, or similar.
• Experience with large‑scale image processing pipelines.
• Experience with synthetic data generation, active learning, weak supervision, or dataset quality improvement.
• Experience with multimodal models, vision‑language models, prompt engineering, OpenAI, or similar AI platforms.
• Experience with cloud ML platforms and production monitoring of ML models.
• Familiarity with mobile applications, warehouse workflows, field operations systems, or image capture workflows.
• Familiarity with Azure DevOps, Git, CI / CD tooling, documentation systems, and practical software delivery processes.
• Experience in electronic components, technology distribution, supply chain, logistics, manufacturing, e‑commerce, or similar B2B environments.

Success in the First 90 Days

Within the first 90 days, the Principal Computer Vision Scientist should be able to:

• Understand the relevant product areas, users, business workflows, image capture workflows, datasets, model use cases, and current technical risks.
• Establish working relationships with Engineering Managers, Team Leads / Architects, Product / Delivery Managers, engineers, QA, DevOps, Data, and business stakeholders.
• Review current Computer Vision and AI / ML work, including models, datasets, evaluation methods, annotation process, production integration, and known quality issues.
• Identify the most important model quality risks, data quality gaps, technical debt items, production risks, and maintainability concerns.
• Define or improve model evaluation criteria, validation process, dataset requirements, and model readiness expectations.
• Help improve technical clarity of the active AI / ML backlog by adding design notes, technical breakdown, dependencies, estimates, and risks.
• Lead at least one meaningful model improvement, prototype, evaluation effort, or production risk‑reduction activity.
• Improve documentation around model architecture, data flows, evaluation results, known limitations, and production behavior.
• Create an initial technical roadmap or remediation plan for Computer Vision work aligned with product priorities and engineering capacity.
• Help onboard or mentor engineers working on Computer Vision, AI / ML, or related product features.

What This Role Does Not Own

This role does not own formal people management for engineers. Engineering Managers remain responsible for hiring, performance management, compensation input, team structure, and capacity planning.

This role does not own business prioritization or user acceptance. Product / Delivery Managers and business stakeholders remain responsible for intake, priority alignment, backlog readiness, UAT coordination, and business acceptance.

This role does not independently commit delivery dates without alignment with Engineering Managers and Product / Delivery Managers.

This role does not own all AI / ML work across the company unless specifically assigned by GEO leadership. The primary responsibility is Computer Vision technical leadership and production‑quality AI / ML implementation for assigned product areas.

This role does not replace Software Architects, DevOps, Data, QA, or Infrastructure ownership. The role will work closely with those teams to make sure Computer Vision solutions are technically sound, production‑ready, and supportable.