Scrabble is partnering with an AI Legal Tech startup.
Responsibilities
Design and deploy production-grade LLM agents capable of operating autonomously for extended periods, including managing state, memory, retries, escalation, and recovery processes.
Build and maintain an integration layer that abstracts the complexities of various legal case management systems (e.g., Clio, Filevine, Litify, MyCase) into a unified, user-friendly interface.
Establish and uphold reliability standards through rigorous evaluations, transcript regression testing, and identification of edge cases.
Refine the conversational abilities of the AI agent, ensuring the voice and tone meet high standards by analyzing transcripts and user interactions.
Develop a streamlined onboarding process for new law firms, accommodating their unique intake forms, processes, and CMS configurations within days.
Collaborate directly with customers, founders, and a small team of developers, taking full ownership of the workstream without reliance on managers or handoffs.
Qualifications
Proven experience shipping LLM-based agents in production environments with real users and handling real failures.
Strong understanding of agent reliability engineering, including writing evaluations, regression tests, and handling edge cases.
Experience normalizing and integrating messy third-party systems, such as legacy APIs and diverse configuration requirements.
Product instincts with a focus on conversational quality; ability to identify and improve conversational missteps.
High standards for engineering quality and reliability, self-motivated to maintain excellence.
Excellent communication skills and ability to work collaboratively in a small team.
Educational qualifications: BTech, BSc, MSc, or equivalent in Computer Science, Engineering, or related fields.
Preferred Skills
Experience working with legal or case management software integrations.
Familiarity with conversational AI, voice interfaces, and natural language processing.
Prior experience in automating workflows within highly fragmented or legacy systems.
Ability to quickly adapt to new tools and environments, with a focus on rapid onboarding and iteration.
Experience
Demonstrated 2+ years of relevant experience deploying LLM agents in production with real-world failure handling.
Experience working with APIs, legacy systems, and multi-system integrations.
Proven ability to handle complex failure scenarios and improve system robustness.
