Forward Deployed Engineer

Prevail Recruiting•San Diego, CA

About The Position

Work directly with customers and internal engineering teams to identify valuable applications for AI, build working solutions, and deploy them into real operating environments. Own the path from understanding a customer’s workflow to integrating data, selecting models, developing applications, and measuring performance after deployment. The role combines software engineering fundamentals with practical knowledge of LLMs, retrieval systems, and AI agents. It requires writing production code, debugging unfamiliar systems, and explaining technical decisions to engineers and business stakeholders. Success means delivering reliable tools that customers adopt and use to improve their work. Responsibilities include translating customer problems into technical solutions; building and deploying AI applications; selecting and integrating models; developing retrieval-augmented generation systems and tool-using workflows; integrating with customer systems; evaluating reliability; engineering production-ready applications; protecting sensitive information; owning deployment and adoption; and turning recurring deployment needs into reusable capabilities and product improvements.

Requirements

Strong software engineering experience, including ownership of applications or services deployed to real users.
Proficiency in Python and experience building APIs, backend services, and data integrations.
Demonstrated experience developing applications with LLMs beyond basic chatbot interfaces.
Practical understanding of prompting, context management, embeddings, retrieval, structured outputs, and tool calling.
Ability to explain model limitations and design evaluation methods that measure task performance.
Experience working with SQL databases, unstructured data, and document processing pipelines.
Familiarity with containers, cloud infrastructure, version control, automated testing, and application monitoring.
Ability to debug problems across data pipelines, model interactions, application code, and infrastructure.
Strong customer communication skills and ability to turn incomplete requirements into a working implementation.
Comfort taking ownership in an early-stage environment with evolving priorities.

Nice To Haves

Previous work as a Forward Deployed Engineer, Applied AI Engineer, Machine Learning Engineer, or customer-facing Software Engineer.
Experience deploying open-weight models, including inference serving, quantization, GPU memory management, and performance optimization.
Experience with fine-tuning or parameter-efficient adaptation, with sound judgment about when these approaches are justified.
Experience building agent workflows with explicit permissions, approval points, retries, and recovery mechanisms.
Experience with multimodal applications involving technical drawings, images, scanned documents, or sensor information.
Experience integrating engineering, manufacturing, enterprise knowledge, or operational software.
Experience with private-cloud, on-premises, edge, or restricted-connectivity deployments.
Experience supporting aerospace, defense, robotics, advanced manufacturing, or other technically demanding customers.
Familiarity with prompt injection defenses, retrieval access controls, and secure handling of model-generated tool requests.
TypeScript or React experience for building usable customer interfaces.

Responsibilities

Work with users to understand workflows, data sources, constraints, and success criteria, and translate customer problems into technical solutions.
Build and deploy AI applications for technical document analysis, knowledge retrieval, workflow automation, engineering support, and decision support.
Evaluate and integrate commercial and open-weight models based on task requirements, accuracy, latency, cost, and deployment constraints.
Build retrieval-augmented generation pipelines for document ingestion, parsing, chunking, embedding, indexing, retrieval, reranking, and source attribution.
Build tool-using AI workflows that connect models to APIs, databases, and software tools using structured outputs, function calling, and controlled execution.
Develop APIs, backend services, and interfaces that connect AI capabilities to existing applications and data.
Create representative test datasets and performance baselines, analyze failures, and improve output quality and reliability.
Implement authentication, access controls, logging, monitoring, error handling, and deployment processes.
Design data boundaries and controls to protect customer information, retrieved content, and model-driven actions.
Troubleshoot deployed systems, gather user feedback, document solutions, and improve systems after launch.
Identify recurring customer needs and translate them into shared components and product improvements.