Director of Machine Learning

About us:

Toronto Interactive is a 14-person team running a niche online store for vintage audio parts. Growth has come from search and word of mouth, not paid ads.

The role:

Toronto Interactive is hiring for the Director of Machine Learning role to take models from notebook to production and keep them working once they're there. Fully remote. You can work from anywhere in Canada. Flexible schedule. We have one team call a day, usually in the morning. We talk mostly in Slack and hop on quick calls when something needs discussing.

In your first few months you will:

• Set up evaluation harnesses and regression tests for model quality
• Orchestrate pipelines with Kubeflow, SageMaker Pipelines, Vertex AI or Airflow
• Fine-tune open-weight models with LoRA when it beats prompting
• Write clean, tested Python and review teammates' code
• Deploy models behind low-latency APIs with FastAPI, Triton, TorchServe or vLLM
• Handle on-call for the ML services the team owns
• Develop RAG pipelines with embeddings, vector stores (pgvector, Pinecone, Weaviate) and rerankers

What we're looking for:

• Familiarity with Docker, Kubernetes and a cloud ML platform (SageMaker, Vertex AI, Azure ML)
• Knowledge of LLM APIs, prompt design and evaluation methods
• BS or MS in computer science, engineering or a related field

Preferred:

• Guardrails, red teaming or safety evaluation for LLM features
• Distributed training with DeepSpeed or FSDP
• CUDA or Triton kernel experience

What you get:

• Flexible hours
• Paid sick days
• Home office allowance
• RRSP matching
• Employee and family assistance program

Only candidates selected for an interview will be contacted.

Start: Immediately.

Demo job created on 11jobs for testing and demonstrations. It is not an actual opening at Toronto Interactive Inc..