Location: Vancouver, British Columbia
About the company:
Our client is a fast-growing, venture-backed B2B SaaS company building an AI-powered competitive intelligence platform that helps revenue teams win more competitive deals. The company is AI-first internally, and its product is trained in part on proprietary data unavailable anywhere else on the open web.
About the role:
We're hiring a Senior Software Engineer to build and optimize state-of-the-art LLM-powered agents at scale, spanning multi-agent orchestration, sub-agent design, and the evaluation frameworks that keep outputs trusted and measurable. You'll own projects end-to-end — architecture, experimentation strategy, and production readiness — while optimizing inference cost, retrieval and query performance, and the feedback loops that help agents improve over time. You'll also help shape the roadmap itself, bringing a technical point of view and working closely with product leadership.
What you'll do:
• Build backend systems for agentic workflows — retrieval pipelines, orchestration layers, and multi-step agent architectures.
• Own evaluation of agentic systems at scale — build and operate frameworks measuring relevance, quality, latency, and task success.
• Design human-in-the-loop systems — feedback mechanisms, review workflows, and correction loops that keep agents accurate over time.
• Optimize LLM workflows end to end — prompt design, retrieval strategy, caching, and latency.
• Collaborate across product, infrastructure, and data teams to shape technical direction and roadmap.
Must Have Skills:
• Production experience with backend systems — search/retrieval, data pipelines, distributed systems, or API-heavy services.
• Experience building and/or evaluating agentic or LLM-powered systems (RAG, multi-step agents).
• Strong Python and software engineering fundamentals (testing, CI/CD, observability).
• Experience with retrieval/search systems and vector databases (PGVector, Pinecone, or Elasticsearch).
• Experience with cloud infrastructure at scale (AWS, GCP, or Azure).
• Regular use of AI coding tools (Copilot, Cursor, Claude Code, or similar).
• Track record of shipping features tied to real user/business outcomes.
• Ability to own a project end-to-end and provide technical direction.
Nice to Have Skills:
• Multi-agent systems or complex orchestration design.
• ML fundamentals (precision/recall, cross-validation, bias-variance tradeoff, overfitting/regularization).
• Open-source contributions in AI/ML.
