Forward Deployed Engineer

About the Role

You will work directly with customers and internal engineering teams to identify valuable applications for AI, build working solutions, and deploy them into real operating environments. You will own the path from understanding a customer’s workflow to integrating data, selecting models, developing applications, and measuring performance after deployment.

The ideal candidate combines strong software engineering fundamentals with practical knowledge of LLMs, retrieval systems, and AI agents. You should be comfortable writing production code, debugging unfamiliar systems, and explaining technical decisions to both engineers and business stakeholders.

Success means delivering reliable tools that customers adopt and use to improve their work.

What You’ll Do

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

Required Qualifications

• 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 the ability to turn incomplete requirements into a working implementation.
• Comfort taking ownership in an early-stage environment with evolving priorities.

Preferred Experience

• Previous work as a Forward Deployed Engineer, Applied AI Engineer, Machine Learning Engineer, or customer-facing Software Engineer.
• Deploying open-weight models, including inference serving, quantization, GPU memory management, and performance optimization.
• Fine-tuning or parameter-efficient adaptation, with sound judgment about when these approaches are justified.
• Building agent workflows with explicit permissions, approval points, retries, and recovery mechanisms.
• Multimodal applications involving technical drawings, images, scanned documents, or sensor information.
• Integrations with engineering, manufacturing, enterprise knowledge, or operational software.
• Deployments in private-cloud, on-premises, edge, or restricted-connectivity environments.
• 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.



What Strong Candidates Can Demonstrate

Candidates should be prepared to walk through an AI system they personally built and deployed, including:

• The user problem and their individual technical contribution.
• Why they selected a particular model and architecture.
• How they connected the system to real data and tools.
• How they measured quality and investigated failures.
• The tradeoffs they made around accuracy, latency, cost, and security.
• Evidence of adoption, time savings, or improved operational performance.