Staff Machine Learning Engineer, Content Mining

Job Location:

Toronto - Canada

Monthly Salary: Not Disclosed

Posted on: 30+ days ago

Vacancies: 1 Vacancy

Job Summary

Content Mining identifies the best sources to acquire content for Pinterest (websites merchants social accounts) optimizes how we acquire it and extracts structured attributes from that content at high scale. Our work powers inspiring accurate and engaging Pins. This is a high performing end to end ML team with a recent paper in KDD: Cross-Domain Web Information Extraction at hiring a Staff ML Engineer to serve as the technical lead for a 5engineer team (4 MLEs you). As a tech lead you will define the multiquarter technical vision and roadmap lead execution and mentor engineers. The majority of your time will be spent with handson designing training and shipping ML systemsespecially LLM/NLP models for extraction and

What youll do:

Technical leadership

Own the longterm architecture roadmap and execution for source discovery acquisition optimization and content understanding.
Lead design reviews set engineering standards and drive crossteam alignment with Product Data and Infra.
Mentor and uplevel MLEs through technical direction pairing and reviews.

Modeling and systems

Train/finetune LLMs and NLP models for classification extraction and instructionfollowing; design eval loops and guardrails.
Design features and frameworks for sharing features across models.
Productionize models for large-scale inference; drive latency reliability and cost efficiency (quantization distillation caching).

Data experimentation and quality

Establish offline/online evaluation gold sets and automated regressions; run A/B and canary/shadow launches.
Work with human and automated labeling sources to define data labeling standards.
Partner on data strategy labeling/weak supervision and feedback loops to expand coverage and improve precision/recall.

Operational excellence

Define and meet SLOs for data quality model performance and serving reliability; lead incident playbooks and postmortems.
Measure and drive downstream impact on revenue and engagement.

What were looking for:

5 years building ML products endtoend including 2 years as a tech lead driving multiquarter roadmaps and crossfunctional execution.
Deep handson experience with NLP/LLM training and inference (PyTorch Python); strong grounding in evaluation prompt/data design and finetuning.
Proven track record shipping models at scale: feature/data pipelines online serving monitoring/observability and cost/perf tradeoffs.
Strong software engineering in Python with an eye for software engineering best practices.
Experience mentoring senior engineers and influencing partner teams.
Masters or PhD in ML related studies.
LLM efficiency techniques (LoRA/adapters distillation quantization prompt caching) and cost control strategies.
MLOps at scale with tools like Airflow Spark/Presto Triton vLLM.

Relocation Statement:

This position is not eligible for relocation assistance. Visit ourPinFlexpage to learn more about our working model.

In-Office Requirement Statement:

We let the type of work you do guide the collaboration style. That means were not always working in an office but we continue to gather for key moments of collaboration and connection.
This role will need to be in the office for in-person collaboration 1-2 times/quarter and therefore can be situated anywhere in the province of Ontario.

#LI-REMOTE
#LI-AK7

Required Experience:

Staff IC