ai engineer for agentic systems

Dwelly is building an AI operating system for residential lettings. Its network of agencies provides real-world data, continuous feedback, and control over complete workflows, supporting AI-driven automation for landlords and tenants. The company operates more than 15,000 properties and $470 million in GMV and is one of the UK’s ten largest lettings operators.

• Design and build core primitives for agentic systems, including memory, context management, tool-calling, orchestration, and feedback loops;
• Build reusable infrastructure that enables new workflows to be introduced quickly with predictable reliability;
• Build an evaluation framework for understanding agent performance and failure causes;
• Establish testing, tracing, debugging, and evaluation of AI systems as part of the engineering process;
• Develop systems for assessing agentic workflows before production deployment;
• Create workflows where coding agents and specialized subagents explore repositories, implement changes, review architecture, check conventions, run evaluations, and iterate with minimal human coordination;
• Use agentic engineering workflows extensively in daily development;
• Design systems where specialized agents, tools, deterministic software, and humans work together effectively;
• Decide when to use LLMs, traditional software, or coordinated agents;
• Work with operational and product teams to understand workflows across acquired agencies;
• Identify high-leverage automation opportunities and turn them into reliable production systems;
• Help define architectural patterns for agentic systems as the company scales.

• Independently deliver complex systems from idea through production;
• Build AI or agentic systems that have operated in production;
• Handle reliability, latency, context management, failure modes, evaluation, and observability challenges;
• Understand agents, tool use, structured outputs, context management, orchestration, evaluation, and feedback loops;
• Use modern LLMs practically to build reliable workflows around probabilistic systems;
• Use coding agents and agentic development workflows in the engineering process;
• Decompose complex engineering tasks across agents or specialized subagents and design mechanisms to verify and improve their outputs;
• Apply strong architectural judgment when choosing between LLMs, deterministic software, human intervention, or combined approaches;
• Have experience with TypeScript / Node.js, or strong experience building AI systems in Python with the ability and willingness to work in a TypeScript-first environment;
• Demonstrate high autonomy, ownership, and comfort making decisions under ambiguity;
• Demonstrate strong product judgment and connect technical decisions to operational and business outcomes;
• Communicate strongly and demonstrate fluency in English;
• Show resilience, adaptability, and the ability to thrive in a fast-paced startup environment;
• Do not primarily seek model research, paper reading, or custom model training;
• Do not rely only on clearly defined backlogs or treat AI primarily as a feature;
• Be interested in reliability work including evaluation, logging, error handling, observability, infrastructure, and operational edge cases.

• Fully remote role;
• Competitive salary;
• Potential equity options based on performance;
• Remote collaboration across the UK, Ireland, and European time zones; candidates should be based within these regions or time zones;
• Residential rentals industry focused on improving the experience of more than 5 million UK rental households and over 100 million households including the EU and US;
• Personal data may be processed and stored for recruitment purposes for up to 365 days.