# Senior Agentic AI Engineer

Full-Time · Senior Washington, DC / Remote (US)

## Mission

You'll architect the autonomous agents that power our intelligence platform-forensic analysis systems that operate across billions of legal relationships to identify the structural vulnerabilities that determine case outcomes. Build the AI that gives litigators an unfair advantage.

### What You'll Build

#### Agentic Workflows

- Build autonomous workflows in n8n that query our Databricks Lakehouse, call LLM services, and validate findings against primary source documents
- Design self-correcting loops to detect hallucinations, dead-ends, and infinite reasoning loops, and recover gracefully
- Turn fragile prompts into robust, observable production systems processing thousands of legal documents daily

#### Intelligence Extraction

- Agents that reconstruct examiner search logs from unstructured USPTO Office Actions
- Systems that map prosecution history amendments across patent families to identify estoppel-creating arguments
- Detection engines that find contradictions between complaint allegations and patent specifications
- Validation agents that verify every finding against exact page and line citations in source documents

#### Production AI Systems

- Deploy and monitor agentic workflows at scale
- Build robust error handling and retry logic for unreliable external data sources (USPTO APIs, PACER)
- Implement versioning, A/B testing, and observability for agent improvements
- Work closely with product, data, and litigation experts to translate legal attack vectors into AI workflows

## Requirements

- 5+ years building production AI/ML systems (not just research prototypes)
- Strong background in workflow orchestration (n8n, Airflow, Temporal, or similar); n8n familiarity is highly valued
- Strong Python engineering skills with production-quality code standards
- Deep understanding of LLM integration, prompt engineering, and RAG architectures
- You treat prompt chains as code-versioned, tested, and reliable

### Bonus

- Experience with Databricks for data retrieval and processing
- Familiarity with vector databases and semantic search systems
- Prior work with legal, medical, or financial document analysis
- Experience designing evaluation harnesses for LLM features

## Compensation

- **Base salary:** $160,000 - $220,000 (based on experience)
- Stock Options
- Performance bonus eligibility