Software Engineer, AI Agents

Quick Facts

• Role: Software Engineer, AI Agents

• Focus: Agentic systems and LLM platform infrastructure for legal AI

Description

Build agentic systems at the core of a legal AI platform, including an orchestration runtime, retrieval-layer reasoning agents, and evaluation infrastructure. Own systems from initial design through production operation, with a strong emphasis on correctness, traceability, confidentiality, and tenant isolation. Establish platform primitives for orchestration, grounded retrieval, and measurable quality that other teams can build on.

Responsibilities

• Design an agentic platform with orchestration of single and parallel sub-agents, durable state/memory, MCP-based tool integration, and trace-level observability

• Build safety-layer agents: tool registration/permissioning, sandboxing, guardrails, output validation, and defenses against prompt injection

• Create multi-step agent workflows with tool use, human-in-the-loop checkpoints, and recovery from partial failures

• Translate model capability and evaluation practice learnings into platform architecture

• Design retrieval-augmented generation with hybrid search over legal corpora and citation grounding to specific sources

• Own and evolve model services for classification, extraction, and embeddings, including custom models with active learning

• Build evaluation datasets, benchmarks, and regression checks for classification, extraction, retrieval, and agent workflows; run automated evaluation in CI and monitor quality/latency/cost

• Establish LLMOps practices: model routing/fallbacks, latency/cost budgets, prompt/context versioning, caching, and quality/reliability monitoring

• Use AI coding agents in development while owning outcomes via specs, test plans, code review, and verification of correctness/security/maintainability

• Build backend services/APIs, write technical design documentation, and collaborate with product/research/security stakeholders

• Participate in on-call rotation to diagnose and resolve incidents affecting model and agent services, and implement durable fixes

Requirements

• 3+ years of professional software engineering experience with backend/distributed/AI depth

• Experience delivering and operating LLM-powered applications in production at scale, including quality evaluation and failure diagnosis

• Experience building tool-using agent workflows or retrieval-augmented generation systems

• Strong production-level proficiency in Python

• Working understanding of the ML lifecycle (evaluation data, versioning, deployment, and monitoring)

• Experience designing evaluation datasets/benchmarks and measuring task completion, citation correctness, policy adherence, latency, and cost

• Experience designing services and APIs consumed by other teams

• Observability tooling experience (Prometheus, Grafana, or OpenTelemetry)

• Strong system design and technical documentation skills

• Excellent communication with domain experts

• Proven day-to-day use of AI coding agents with sound judgment

Benefits

• Compensation range: $180,000–$260,000 USD annually; may be eligible for an annual bonus