AI Engineer

We're hiring an AI Engineer to own the intelligence layer of our platform. You'll design and build the extraction pipelines, model orchestration, prompt architectures, and evaluation frameworks that turn messy insurance documents into structured, underwriting-ready data. This isn't a research role. You'll be shipping production AI systems that process real broker submissions, extract data from ACORD forms and policy documents, and feed structured outputs into downstream underwriting workflows. You need to be comfortable with the full lifecycle—from prompt design to deployment to monitoring to iteration based on production feedback.

Responsibilities:

• Design, build, and maintain production LLM extraction pipelines that process insurance documents (ACORD 125/140, loss runs, supplemental apps, SOVs, SPAs) into structured data.
• Own the multi-pass extraction architecture end-to-end: OCR layer, LOB detection, LOB-specific structured extraction, confidence scoring, and fallback logic.
• Write, version, and optimize extraction prompts across insurance lines of business: General Liability, Commercial Property, Workers' Comp, Professional Indemnity, W& I, and more.
• Build and maintain evaluation frameworks: ground truth datasets, extraction accuracy benchmarking, regression testing, and automated quality scoring.
• Design feedback learning loops that improve extraction quality over time: correction feedback from underwriters, document template recognition, and triage calibration.
• Implement model routing and fallback strategies: cost optimization, latency management, provider failover, and model selection logic.
• Build document classification and triage systems that automatically identify document types, LOBs, and routing priorities from incoming submissions.
• Own LLM observability: prompt tracing, cost tracking, latency monitoring, A/B testing prompt variants, and debugging extraction failures.
• Collaborate with product and engineering to define extraction schemas, field mappings, and output contracts that downstream systems consume.
• Evaluate and integrate new models, techniques, and tools as the LLM landscape evolves: vision models, structured output modes, fine-tuning strategies, and agentic patterns.

Requirements:

• 4+ years of software engineering experience with strong Python proficiency and at least 2 years of hands-on work building AI/ML-powered production systems.
• Deep, practical experience with LLM APIs in production: prompt engineering, output parsing, structured extraction, token management, error handling, and retry strategies.
• Experience building document understanding or information extraction pipelines: OCR, PDF processing, form extraction, or NLP-based data extraction at scale.
• Strong understanding of evaluation methodology: you know how to measure extraction quality, build ground truth datasets, and set up automated regression testing for AI outputs.
• Hands-on experience with at least one LLM observability or evaluation platform (Langfuse, LangSmith, Braintrust, Weights and Biases, or equivalent).
• Production experience with AWS or equivalent cloud platform: you've deployed and monitored AI pipelines, not just runned notebooks.
• Daily, habitual use of AI tools in your own engineering workflow: Claude Code, Copilot, Cursor, or equivalent for coding, debugging, and research.
• Ownership mindset: you don't wait for someone to hand you a perfectly scoped ticket; you identify extraction quality issues, propose solutions, build them, and measure the impact.
• Clear communication skills: you can explain model tradeoffs, extraction accuracy metrics, and pipeline architecture to product managers and engineers who aren't AI specialists.

Core Tech Stack:

• Python: primary language for all AI/ML pipelines, extraction logic, and evaluation scripts.
• LLM APIs: Claude (Anthropic), Gemini (Google).
• Multi-pass extraction architecture.
• Workflow orchestration for multi-step extraction and submission processing pipelines.
• Structured storage and vector embeddings for document similarity and template matching.
• LLM observability, tracing, prompt versioning, and evaluation.
• AWS: compute, storage, deployment (S3, Lambda, SQS, ECS).

Skills

AWS, ECS, LLM, Lambda, Large Language Model, Python, S3, Sqs