About the role

Software Engineer, AI at Klue, building LLM-powered agents for workflow automation in Toronto. Collaborating on search retrieval optimization and ML-powered solutions.

Responsibilities

Design and implement retrieval-augmented generation (RAG) systems with agentic workflows
Build and optimize retrieval pipelines using BM25, dense retrieval, hybrid retrieval, and re-ranking approaches
Develop evaluation pipelines for retrieval and generation using offline metrics and human-in-the-loop evaluations
Experiment with query rewriting, expansion, and classification to improve retrieval relevance
Collaborate closely with Product to bring ML-powered search agents into production
Profile, debug, and optimize latency, accuracy, and scalability of retrieval and generation components
Contribute to the design of data pipelines for training retrieval and ranking models

5+ years of software engineering experience
Experience with information retrieval systems, search relevance, and ranking models
Expertise in Python, with experience in frameworks such as PyTorch, TensorFlow, or JAX
Familiarity with LLMs, prompt engineering, and retrieval-augmented generation pipelines
Understanding of evaluation methods for search systems including offline metrics and user-facing evaluation
Experience with vector database infrastructure and traditional search engines
Understanding of data pipelines and large-scale data handling
Ability to work independently and collaboratively in a fast-paced environment
Develop and implement CI/CD pipelines
Knowledge of query understanding, document summarization, and other content enrichment strategies
Expertise in automated LLM evaluation