Senior AI Engineer
Sweet James
- Location
- Onsite (Newport Beach, California)
- Employment
- Full-time
- Level
- Senior Level
About the role
Sweet James is a rapidly growing, AI-powered personal injury law firm seeking a Senior AI Engineer to build and own end-to-end AI features. The role involves developing production systems for document processing, agent automation, and integration with legal tech stacks to drive efficiency.
Skills
Full job details
Sweet James Accident Attorneys is a rapidly growing, AI-powered personal injury firm based in Orange County, CA. In less than a decade, we've built a nationally recognized brand and a proprietary AI and automation stack that make us one of the largest and most efficient providers of legal services in the country. Our growth runs on a culture of excellence, compassion, and results: Our team is committed to exceptional client experience and to supporting one another in a fast-paced, collaborative environment. We invest in our people, encourage professional growth, and provide meaningful opportunities to build rewarding careers while making a real impact in our clients' lives.
The role
Technology is how we practice law, not a department beside it. We have built a highly automated, AI-forward experience for every client-facing role, from intake to settlement, and it is one of the leading examples of applied AI in the legal field: Our case teams run at a fraction of the headcount our largest peers need for the same work. The systems behind that result are built in-house, and you will be one of the people building them.
Sweet James is seeking a Senior AI Engineer to own AI features end to end: Learn the work from the people who do it, build the system, measure it, ship it, and run it in production. The systems you will work on read and file documents, pull facts from medical records and police reports, answer questions across a case's full history, and run agents that chase records and draft routine work for staff to approve.
You will report to the Lead AI Engineer, who sets the architecture and build plan; you own delivery within it. This is an individual-contributor seat for someone who drives work forward with urgency and high quality without waiting to be told how, and whose work reaches case teams within weeks, not quarters.
Responsibilities
• Build: Ship production AI features, including structured extraction, classification and routing, case search, and agents that act with a person approving the result.
• Integrate: Connect those features to the firm's systems (e.g., Salesforce/Litify, Microsoft Graph for Outlook and SharePoint, Zoom, Twilio, Retell, DocuSign).
• Measure: Build the evaluation sets and tests for your features so every model or prompt change is measured on facts captured, facts missed, wrong answers, cost, and speed before it ships.
• Run: Deploy your services as containers on Azure, monitor them, and fix them when they break.
• Learn and train: Sit with intake, case management, records, and litigation staff before automating their work, and train them on what you ship.
• Document: Write each system down so anyone on the team can run it.
Qualifications
Required:
• Experience: 5+ years building production software, including 2+ years shipping LLM-powered features to real users.
• Ownership: A record of taking ambiguous problems to shipped outcomes with little supervision, and of finishing what you start.
• LLM craft: Hands-on work with LLM APIs (e.g., Anthropic, OpenAI) spanning prompt design, tool use, structured outputs, retrieval, and evaluation.
• Stack: Strong Python or TypeScript/Node (both preferred), SQL and Postgres, REST APIs and webhooks, Git, Docker, and cloud deployment (Azure preferred).
• Communication: Care with sensitive data and the ability to explain technical work in plain language.
Preferred:
• Models and tooling: Vector search (e.g., pgvector), evaluation frameworks, and agent frameworks or the Model Context Protocol.
• Workflow automation: Platforms such as n8n.
• Salesforce: APIs and SOQL; Litify a plus.
• Voice: Voice AI or telephony (e.g., Retell, Twilio, Zoom).
• Documents and industry: OCR and document processing; legal, insurance, or healthcare data