Skip to main content
TALENT247 // SPECIALIST AI EXECUTIVE SEARCH & TECHNICAL RECRUITMENT
ML · Machine Learning

Staff Machine Learning Engineer, Multilingual Embeddings & Reranking

Toronto, ON, Canada · Full-time · Hybrid

AI Specialism

Machine Learning

Workplace

Hybrid

Location

Toronto, ON, Canada

Compensation

CA$195k – CA$270k

annual package

Role Dossier & Specification

REF: STAFF-MACHINE-LEAR

THE CLIENT & MANDATE

Our client is a multi-billion-dollar enterprise AI foundation model company headquartered in Toronto whose multilingual language models, dense/sparse embeddings, and neural rerankers power production search and agentic RAG workloads across the Global 2000. Unlike general consumer chat labs, their entire research roadmap is dedicated to enterprise-grade accuracy, cross-lingual retrieval across 100+ languages, verifiable citations, and high-efficiency enterprise deployment.

They have retained Talent247 to appoint a Staff Machine Learning Engineer, Multilingual Embeddings & Reranking in Toronto. Sitting as a technical pillar within their Retrieval & Representation Learning group, you will lead the architecture, contrastive pre-training, and post-training distillation of next-generation text and multimodal embedding and cross-encoder reranking models used by millions of enterprise knowledge workers daily.

KEY RESPONSIBILITIES

  • Set technical direction and own end-to-end training recipes for frontier multilingual text and multimodal (slide, table, chart, PDF) embedding and late-interaction/cross-encoder reranking models.
  • Design large-scale contrastive pre-training, synthetic query-document generation, LLM-guided hard-negative mining, and distillation pipelines across 100+ high- and low-resource languages.
  • Pioneer Matryoshka representation learning, binary/int8 vector quantization, and asymmetric query-document encoding architectures that slash customer vector-index RAM costs by 8x–32x without recall degradation.
  • Partner with enterprise agent and RAG research teams to co-train retrieval models alongside reasoning LLMs for multi-hop query decomposition and tool discovery.
  • Build gold-standard enterprise retrieval benchmarks spanning finance, legal, code, technical manuals, and noisy OCR documents, driving measurable gains beyond public academic leaderboards.
  • Mentor senior ML engineers and collaborate with inference teams to ship sub-15ms p99 reranking endpoints in production.

TECHNICAL ENVIRONMENT & OFFERING

Their modeling stack runs on JAX and PyTorch across large multi-slice TPU and H100 GPU clusters, backed by petabyte-scale multilingual data curation pipelines. Based hybrid in downtown Toronto, this Staff mandate offers CA$195,000–CA$270,000 base salary plus substantial equity, premier healthcare benefits, and full relocation/immigration support.

Core Requirements

ESSENTIAL QUALIFICATIONS

  • 7+ years of hands-on experience in machine learning engineering or applied research, with deep specialization in neural information retrieval, representation learning, or foundation model training.
  • Expert proficiency in PyTorch or JAX, distributed multi-node training, contrastive loss formulations, knowledge distillation, and large-scale synthetic data curation.
  • Proven track record shipping state-of-the-art dense/sparse embedding models, cross-encoder rerankers, or multimodal retrieval systems at scale.
  • Deep familiarity with ANN vector search internals (HNSW, DiskANN, PQ, binary/scalar quantization) and storage-vs-recall tradeoffs in enterprise RAG.
  • Track record of technical leadership, experimental rigor, and mentoring engineers in a fast-paced AI lab environment.

PREFERRED / BONUS EXPERIENCE

  • Publications at SIGIR, ACL, EMNLP, NeurIPS, ICML, or ICLR, or top-ranked releases on major retrieval/embedding benchmarks (MTEB, BEIR, MIRACL, ViDoRe).
  • Experience with multimodal vision-language document retrieval or agentic multi-hop search loops.

Position Telemetry

Location Toronto, ON, Canada (Hybrid)
Department Machine Learning
Employment type Full-time
Workplace type Hybrid
Salary range
CA$195k – CA$270k annual package