AI Solutions Engineer
Southern California · Applied AI and process improvement
What we’re looking for
- Hands-on experience building AI-assisted applications, automations or data workflows used by real teams.
- Strong practical knowledge of APIs, structured data, retrieval systems, evaluation and prompt design.
- Ability to work with imperfect business data and turn ambiguous needs into testable solutions.
- Clear documentation habits and a thoughtful approach to privacy, access controls and human review.
Preferred qualifications
- Experience with ecommerce, product information, CRM, customer support or service operations.
- Familiarity with RAG, vector search, workflow platforms and lightweight web application development.
- Ability to explain technical tradeoffs clearly to non-technical colleagues.
What success looks like
Reliable tools that employees actually use, measurable time savings, documented limitations and a maintainable path from pilot to daily operation.
Start your application
Tell us a little about yourself. Our team will review your information and follow up if there is a potential fit.
The opportunity
This is an applied, hands-on role for someone who wants to turn fragmented knowledge and repetitive work into reliable AI-assisted systems. You will begin by learning Arthur Imaging’s product, customer and operational data, then build internal knowledge tools and automations that save measurable time. The goal is not AI demos or pure research; it is useful workflows that are grounded in company information, tested against real examples and safe enough for daily use.
Key responsibilities
- Map existing workflows with sales, ecommerce, product-data, service and operations teams; convert vague requests into clear and testable specifications.
- Audit, clean and organize the source data required for reliable AI workflows, including product information, service knowledge and internal documentation.
- Design retrieval-based knowledge tools that answer questions from approved company sources and provide traceable references.
- Build automations and lightweight integrations using APIs, webhooks, scripts or workflow platforms to reduce repetitive manual work.
- Create evaluation sets and acceptance criteria; measure accuracy, failure cases, time saved and operational value before release.
- Design safeguards for hallucination, prompt injection, sensitive information, access controls and human review.
- Monitor model usage, latency and cost; select the simplest technical approach that meets the quality requirement.
- Document architecture, prompts, data sources, limitations, maintenance procedures and user instructions.
- Pilot customer-facing AI workflows only after internal testing, with clearly defined scope, ownership and escalation paths.