AI Engineer Role

Peregrine Advisors LLC; Washington (District of Columbia); On-site; USD 120,000 - 190,000; Full time

Job summary

Peregrine Advisors LLC is seeking AI Engineers to design and operate systems that turn data, documents, models, and AI services into useful applications and workflows. The work spans from problem understanding to deployment, monitoring, incident response, and continuous improvement.

The role emphasizes connecting data pipelines, models, APIs, user experiences, and business rules, then testing the complete system with proper security, traceability, and oversight.

Responsibilities

• Document and content pipelines that apply OCR, parse layouts, extract and validate information.
• Predictive, classification, or retrieval services combining models, pretrained models, externally provided AI services, and deterministic software.
• Search and knowledge applications using retrieval, retrieval-augmented generation (RAG), and traceable references.

Job description

The work

AI Engineers build and operate systems that turn data, documents, models, and AI services into useful applications and workflows. The work runs from understanding the problem and preparing inputs through model or service selection, system design, evaluation, deployment, monitoring, incident response, and continuous improvement.

Description

AI Engineers build and operate systems that turn data, documents, models, and AI services into useful applications and workflows. The work runs from understanding the problem and preparing inputs through model or service selection, system design, evaluation, deployment, monitoring, incident response, and continuous improvement. This is systems engineering around AI behavior. AI Engineers connect data pipelines, models, APIs, user experiences, business rules, and human-review paths, then test the complete system under conditions that reflect how people will use it. They make limitations visible and match security, traceability, and oversight to the consequences of the work.

What You'll Build

• Document and content pipelines that apply OCR, parse layouts, extract and validate information, preserve provenance, and make authoritative content usable downstream.
• Predictive, classification, recommendation, natural language, computer vision, or decision-support services that combine custom models, pretrained models, externally provided AI services, and deterministic software.
• Search and knowledge applications that use retrieval, retrieval-augmented generation (RAG), controlled source access, and traceable references.
• Workflow and agent-based systems with scoped tools, permission boundaries, approval paths, exception handling, and audit trails where agentic patterns fit the problem.
• Evaluation and operations capabilities that version, test, release, trace, monitor, diagnose, and roll back models, prompts, configurations, retrieval indexes, and supporting software.

Who You Are

You see the model as one part of a larger system. You can move between user needs, data, software, model behavior, and production operations without losing sight of the outcome, and you choose the simplest approach that meets the need.

You test assumptions against evidence, communicate tradeoffs clearly, and work comfortably with domain experts, users, data engineers, data scientists, Machine Learning Engineers, security specialists, governance teams, and operations teams. You know when to build, when to integrate, when to accelerate, and when AI is not the right answer.

What You Bring

• A working foundation in software engineering, including programming, APIs, testing, version control, and the design of services or applications.
• Practical experience with structured or unstructured data, including preparation, validation, metadata, access controls, quality, and traceability.
• Applied understanding of AI and machine learning, including model or service selection, evaluation design, error analysis, and clear communication of limitations.
• Experience contributing to production reliability through deployment discipline, observability, performance and cost management, security, incident response, or related operational practices.
• The judgment to match evaluation, documentation, safeguards, and human oversight to the context and consequences of the system.

What Openings May Require

An opening may emphasize document intelligence, conventional ML, generative AI, RAG, agents, computer vision, AI platforms, machine learning operations (MLOps), generative AI operations, evaluation, security, governance, or a particular mission domain.

Specific openings may name programming languages, cloud environments, model providers, ML frameworks, document-processing tools, search or vector platforms, workflow and agent frameworks, data stores, container and deployment platforms, observability systems, or regulated-development practices. OPEN Data Jobs will identify the required and preferred capabilities with each opening.

Compensation And Benefits

Compensation, benefits, work location, and employment terms are set for each specific opening and will be stated with that opening.