About the job
Join Klue: Pioneers in Competitive Intelligence
At Klue, we are transforming the future of competitive intelligence. Our mission is to empower our customers with actionable insights derived from vast amounts of competitive data.
Join Our Engineering Team!
We are on the lookout for a Software Engineer with production experience to become a vital part of our AI engineering team in Toronto. In this role, you will be instrumental in developing and enhancing LLM-powered agent systems, converting millions of data points, including news, press releases, and CRM data, into strategic insights for our clients.
This position is designed for hands-on builders. From day one, you will be shipping code into our agent architecture, collaborating with senior engineers, managing well-defined segments of larger projects, and progressively taking on greater responsibilities. If you have previously deployed a backend system and are eager to delve into retrieval, RAG, and agent infrastructure, we would love to connect with you.
Your Contributions
As part of our core AI services, you will work on:
- Ambient Agents of Klue: Our agent runtime that includes scheduled and event-driven agents. You'll have the opportunity to enhance existing workflows, build new scheduled agents, or refine tooling.
- Multi-agent Orchestrator: This service manages agents that exclusively communicate with data through MCP servers. Tasks may include integrating new MCP tools, refining prompts and streaming behaviors, or evaluating chat quality.
- Search and Retrieval Layer: Overseeing our query layer for Klue's data, you may add new indexes, enhance query endpoints, or improve retrieval chunking strategies.
During your first six months, you will be expected to deliver small, manageable tasks under guidance, take full ownership of a feature, and gradually contribute to larger initiatives.
Your Role
- Deliver Backend Code: Develop and refine agent services, workflows, and MCP tools.
- Enhance Retrieval and Ranking: Collaborate with senior engineers on hybrid retrieval methods, query rewriting, and ranking strategies. Bring your curiosity about BM25 vs. dense vs. hybrid retrieval, and we will guide you through the rest.
- Conduct Evaluations: Design and execute evaluations to assess agent performance and functionality.
