Software Engineer

We are looking for a hands-on AI Engineer who has experience building and deploying production-grade LLM agents. You will own an end-to-end agentic workflow, including architecture, integrations, memory, reliability, evaluation and customer-specific customisations.

Responsibilities:

• Design, build and deploy autonomous LLM agents capable of running for extended periods with reliable state management, memory, retries and recovery.
• Develop integrations with legal case management platforms such as Clio, Filevine, Litify and MyCase.
• Build scalable abstraction layers to handle third-party APIs and customer-specific workflows.
• Implement agent evaluation frameworks, transcript regression testing and reliability monitoring.
• Optimise agent conversations, voice interactions and decision-making capabilities.
• Work closely with customers and founders to understand workflows and deliver customised agent solutions.
• Take complete ownership of architecture, development, integrations and production performance.

Requirements:

• Strong hands-on experience building and deploying production-grade LLM agents.
• Experience with Agentic AI, LLMs, tool calling and multi-step autonomous workflows.
• Strong understanding of agent memory, state management, retries, orchestration and error recovery.
• Experience building integrations with third-party APIs and complex enterprise systems.
• Experience with LLM evaluation, testing frameworks, transcript analysis or trace-based regression testing.
• Strong software engineering and problem-solving skills.
• Product mindset with the ability to independently own and deliver end-to-end solutions.

Preferred Experience:

• Experience with LangGraph, LangChain or similar agent orchestration frameworks.
• Exposure to voice AI, conversational agents and speech-based workflows.
• Experience integrating with CRM, CMS or case management platforms.
• Experience building reliable AI systems for real-world production environments.

Ideal Candidate:

• An AI Engineer who has built and shipped real-world autonomous LLM agents in production, understands agent reliability and evaluation, and can independently own complex AI workflows end-to-end.