Applied AI Engineer

Jobtailor; California (MO); On-site; USD 140,000 - 180,000; Full time

Job summary

Jobtailor is seeking a Generation Lead for Document Intelligence to design prompts, build context windows, and shape per-entity schemas for reliable AI-generated outputs.

You will own evaluation datasets, monitor quality, and collaborate with the Engineering and Processing teams to deliver production-ready GenAI features. This role emphasizes cost-aware model selection and multimodal understanding.

Qualifications

• Strong GenAI craft with prompt engineering and structured outputs.
• Experience building datasets and rubrics; measure quality.
• Experience shipping LLM features into production.
• Ability to connect model performance to product impact.
• Experience with multimodal inputs and OCR data.
• Familiarity with eval platforms is a plus.

Responsibilities

• Own the Generation half of Document Intelligence.
• Design and iterate system prompts, few-shot prompts, and context assembly for generation recipes.
• Define per-entity target schemas and domain validators.
• Build generation-quality evaluation datasets and rubrics, offline and online, using eval pipeline.
• Monitor and address quality regressions.
• Assemble multimodal context windows from keyframes, transcripts, and OCR.
• Choose models and token budgets based on quality, cost, and latency tradeoffs.
• Use LLMX for model access and Attachments for ingestion.
• Hand off validated entities to domain-owning teams.
• Report to the Engineering Lead and collaborate with the Processing side of Document Intelligence.

Skills

GenAI Craft

Prompt Engineering

Dataset Building

Quality Evaluation

Model Performance

Multimodal Handling

RAG & Retrieval

LLM Features

Tools

LLMX

Eval Pipeline

OCR

Retrieval Systems

Job description

• Own the Generation half of Document Intelligence
• Design and iterate system prompts, few-shot prompts, and context assembly for generation recipes
• Define per-entity target schemas and domain validators
• Build generation-quality evaluation datasets and rubrics, offline and online, using LLMX’s eval pipeline
• Monitor and address quality regressions
• Assemble multimodal context windows from keyframes, transcripts, and OCR
• Choose models and token budgets based on quality, cost, and latency tradeoffs
• Use LLMX for model access and Attachments for ingestion
• Hand off validated entities to domain-owning teams
• Report to the Engineering Lead and collaborate with the Processing side of Document Intelligence

Requirements

• Strong applied GenAI craft, including prompt engineering, structured output/tool-use, RAG, and retrieval-context patterns
• Experience building datasets and rubrics and measuring factuality, relevance, and quality
• Experience shipping LLM features into production services
• Ability to connect model performance to product impact
• Comfort working with multimodal inputs including video, PDF, audio, and images converted to text or structured output
• LLM observability/cost awareness or experience with an eval platform (nice to have)
• Document, PDF, or video understanding; OCR; retrieval systems (nice to have)
• Light fine-tuning experience or familiarity with the industrial/maintenance domain (nice to have)
• Legally authorized to work in the United States

Core Competencies

Demonstrates expertise in Generative AI, including prompt engineering and dataset creation, while effectively connecting model performance to product impact. Proficient in handling multimodal inputs and ensuring quality through evaluation datasets and rubrics.

Highest-signal resume keywords

• Generative AI Craft
• Prompt Engineering
• Dataset Building
• LLM Feature Production
• Multimodal Input Handling

ATS Optimization Keywords

Hard Skills

• Prompt Engineering
• Dataset Building
• Quality Evaluation
• Model Performance Measurement
• Light Fine-Tuning

Industry Keywords

• Document Intelligence
• Multimodal Context
• Structured Output
• Video Understanding
• PDF Understanding

Tools & Technologies

• LLMX
• Eval Pipeline
• OCR
• Retrieval Systems