The Role
Design, train, and deploy ML models for fraud detection, synthetic identity, and deepfake detection. Build evaluation pipelines, labeled datasets, benchmarks, and monitoring. Productionize models with backend engineering for latency, throughput, and reliability. Research adversarial approaches, multimodal fusion, and active-learning; partner with product and threat intelligence to convert fraud patterns into trainable signals.
Summary Generated by Built In
About the Role
We're hiring a Machine Learning Engineer to join our core ML team in Toronto. You'll design, train, and deploy the models that power Tofu's fraud detection, deepfake analysis, and identity verification systems — directly shaping the accuracy and reliability of the product at scale.
What You'll Do
- Design, train, and ship ML models for fraud detection, synthetic identity classification, and deepfake (audio, image, video) detection.
- Build and maintain robust evaluation pipelines, including labeled datasets, benchmarks, and continuous monitoring for model drift.
- Productionize models in collaboration with backend engineers — owning latency, throughput, and reliability requirements end-to-end.
- Research and prototype novel approaches to adversarial fraud, including multi-modal signal fusion and active-learning loops.
- Partner with product and threat intelligence to translate emerging fraud patterns into trainable signals.
What You'll Bring
- A bias toward shipping — comfortable balancing research rigor with pragmatic delivery in a fast-paced environment.
- Strong analytical and problem-solving skills, with deep curiosity about adversarial systems.
- Excellent communication and the ability to explain trade-offs and model behavior to non-ML stakeholders.
- A collaborative mindset and willingness to mentor more junior engineers and researchers.
- Strong technical documentation skills.
Required Experience
- Bachelor's degree in Computer Science, Machine Learning, Statistics, or a related field (Master's or PhD a plus).
- 5+ years of professional experience building and deploying ML systems in production.
- Expert proficiency in Python and modern ML frameworks (PyTorch, TensorFlow, JAX).
- Hands-on experience with at least one of: computer vision, audio/speech models, NLP, or anomaly detection.
- Experience deploying models on cloud platforms (AWS, GCP, or Azure) using Docker and Kubernetes.
- Strong SQL and data wrangling skills.
- Experience with vector databases, embeddings, and large-scale retrieval (Elasticsearch, FAISS, pgvector).
Preferred Experience
- Experience with deepfake detection, biometrics, or generative model forensics.
- Experience with MLOps tooling (MLflow, Weights & Biases, Kubeflow, SageMaker).
- Experience in fraud, trust & safety, or security-adjacent domains.
- Familiarity with adversarial ML and red-teaming techniques.
Why Tofu and this Role
- You'll help build the trust layer for the internet — one of the defining problems of the AI era.
- You'll join a team obsessed with building a generational company.
- Early engineers have real ownership, real impact, and unlimited growth.
Benefits & Perks
- Competitive salary + meaningful equity.
- Comprehensive health benefits.
- 3 weeks of vacation.
- New laptop and gear to do your best work.
- Tofu swag your friends will want to steal.
Skills Required
- Bachelor's degree in Computer Science, Machine Learning, Statistics, or related field
- Master's or PhD (a plus)
- 5+ years professional experience building and deploying ML systems in production
- Expert proficiency in Python
- Expert proficiency in modern ML frameworks (PyTorch, TensorFlow, JAX)
- Hands-on experience with at least one of: computer vision, audio/speech models, NLP, or anomaly detection
- Experience deploying models on cloud platforms (AWS, GCP, or Azure) using Docker and Kubernetes
- Strong SQL and data wrangling skills
- Experience with vector databases, embeddings, and large-scale retrieval (Elasticsearch, FAISS, pgvector)
- Bias toward shipping; strong analytical and problem-solving skills
- Excellent communication, mentoring ability, and technical documentation skills
- Experience with deepfake detection, biometrics, or generative model forensics
- Experience with MLOps tools (MLflow, Weights & Biases, Kubeflow, SageMaker)
- Experience in fraud, trust & safety, or security-adjacent domains
- Familiarity with adversarial ML and red-teaming techniques
Am I A Good Fit?
Get Personalized Job Insights.
Our AI-powered fit analysis compares your resume with a job listing so you know if your skills & experience align.
Success! Refresh the page to see how your skills align with this role.
The Company
What We Do
Tofu is a specialized fraud detection company focusing on application-stage fraud to protect organizations from security breaches and compliance risks. By leveraging machine learning trained on billions of data points, Tofu analyzes millions of applicants with high accuracy to flag suspicious activity, enabling recruiting teams to focus their time and efforts on building relationships with legitimate candidates.
