Senior Software Engineer, Applied AI

• Lead the design and implementation of multi-step agentic pipelines, LLM-based applications, and AI-powered workflows, taking end-to-end responsibility from architecture through to production. 
• Define and own evaluation frameworks and feedback loops for AI systems in production: prompt regression suites, output quality monitoring, and continuous improvement cycles. 
• Make architectural decisions on AI system design (model selection, orchestration strategy, retrieval approach, latency and cost trade-offs) and explain the reasoning clearly to engineering and product stakeholders. 
• Identify high-impact AI opportunities in complex client environments, translate them into actionable technical proposals, and lead their execution. 
• Integrate modern AI capabilities (agentic orchestration, structured generation, tool use, model routing) into existing architectures with a clear eye on reliability and maintainability. 
• Set the bar for engineering practice on the team: code quality, observability, evaluation rigour, and responsible AI deployment. 
• Use AI coding tools (Claude Code, Cursor, Copilot, or similar) as a primary part of your development workflow, not an occasional aid, and help the team level up in how they work. 
• Stay at the frontier of the AI tooling ecosystem and bring informed opinions on what's worth adopting into practice.

Key Requirements

• 5+ years of experience building production software systems, with at least 2 years working on AI or ML-driven features in production.
• Deep understanding of modern AI concepts: LLMs, embeddings, retrieval systems, agentic design patterns, evaluation frameworks, and observability for AI systems.
• Strong engineering fundamentals (Python and/or TypeScript/Node.js) and a track record of building reliable, scalable backend services.
• Demonstrated ability to lead technical decisions (architecture, trade-offs, technology selection) and communicate them clearly across engineering and product.
• Experience designing and owning systems under real-world constraints: performance, cost, reliability, and long-term maintainability.
• Ability to take ambiguous, open-ended problems and drive them to well-scoped, shippable solutions with minimal direction.
• High agency: you operate autonomously, make decisions, and flag clearly when you need input.
• Defaults to asking "can an agent or model handle this?" before reaching for a traditional approach, and knows when the answer is no.
• Pragmatic over perfectionist: you know when "good enough in production" beats "perfect in theory," and when it doesn't.
• Low ego, high standards: you hold the bar without holding the room hostage.

Nice-to-Have

• Has shipped real products using AI-assisted coding workflows and can speak concretely to what changed in their engineering process.
• Has designed and operated LLM evaluation harnesses, red-teaming pipelines, or output regression frameworks at scale.
• Active user of (or contributor to) the emerging AI tooling ecosystem: MCP, agent protocols, model routers, LLMOps tooling.
• Experience with the full prompt engineering lifecycle in production: versioning, A/B testing outputs, and monitoring for drift.
• Has rebuilt an internal workflow or tool they previously did manually by replacing it with an agentic system, and can describe what they learned.
• Comfortable reading frontier AI research and forming opinions on what's worth adopting versus what's hype.
• Experience mentoring or technically guiding other engineers on AI system design.

* The technical interview focuses on real problems, not algorithmic puzzles. We evaluate candidates holistically. 

If you don't meet every requirement listed above, apply anyway. We care more about how you think and work than whether your CV matches a checklist.