Application Engineer – LLM

Salt Digital Recruitment; United States; On-site; USD 120,000 - 180,000; Full time

Job summary

Salt Digital Recruitment seeks an LLM Application Engineer to bridge advanced model capabilities with production-ready products. You will design agent workflows, integrate models and tools, and continually improve AI behavior in a full-stack engineering role.

You will drive end-to-end implementations from requirements to deployment, focusing on reliability, latency, and cost while collaborating with product and engineering teams.

Qualifications

• Strong software engineering fundamentals with production applications.
• Hands-on experience developing LLM, Generative AI or agent-based systems.
• Experience building AI-powered apps beyond basic model integrations.
• Design prompts, agent workflows, evaluations and AI behaviour.
• Understanding of tool calling, context engineering, memory and multi-step agent execution.
• Ability to write clean, maintainable, production-ready code.

Responsibilities

• Build and ship LLM-powered applications and AI agent workflows.
• Design systems supporting reasoning, memory, tool use and multi-step execution.
• Build orchestration pipelines transforming model outputs into safe actions.
• Integrate LLMs with APIs, databases, search and internal/external tools.
• Develop prompting, context engineering and structured outputs for reliability.
• Create evaluation frameworks and datasets to measure AI quality and regressions.
• Debug AI systems across model behavior to product experience.

Skills

Software engineering fundamentals

LLM/Generative AI

Agent-based systems

Prompts and prompting

Agent workflows design

Evaluations and AI behaviour

Tool-calling and multi-step execution

Production-quality code

System-to-product translation

Tools

Python

LLM APIs

Vector databases

Retrieval systems / RAG

PyTorch

JAX

Backend services

Job description

About the Opportunity

We are partnering with a fast-growing technology company developing a new generation of AI-native applications designed to make everyday tasks, communication, organization and workflows more intelligent and intuitive. The team is building proactive AI experiences with a strong focus on persistent context, reliable long-running workflows and successful real-world task completion. They are looking for an LLM Application Engineer to build the intelligence layer behind these experiences, turning advanced model capabilities into reliable, scalable and intuitive products.

About the Role

As an LLM Application Engineer, you will work at the intersection of Large Language Models, software engineering and product development. You will design agentic workflows, improve model behavior and build the systems required to transform probabilistic AI outputs into dependable user experiences. This is an end-to-end engineering role.

You will work from understanding user and product requirements through to designing agent workflows, integrating models and tools, building evaluation systems and continuously improving AI behavior in production.

What You'll Be Doing

• Build and ship LLM-powered applications and AI agent workflows.
• Design systems supporting reasoning, planning, memory, tool use and multi-step execution.
• Build reliable orchestration pipelines that transform probabilistic model outputs into predictable, observable and safe actions.
• Integrate LLMs with APIs, databases, search capabilities, internal services and external tools.
• Develop prompting, context engineering, structured outputs and tool-calling approaches to improve model behavior.
• Build evaluation frameworks and datasets to measure AI quality, reliability and regressions.
• Debug AI systems across the entire stack, from model behavior and prompts through to orchestration, backend services and product experience.
• Optimize AI systems for quality, latency, scalability and cost.
• Establish production practices across observability, tracing, experimentation, evaluation and continuous improvement.
• Work closely with product and engineering teams to translate ambiguous product challenges into working AI solutions.
• Continuously improve AI workflows using real-world usage, evaluation results and production performance.

What We're Looking For

• Strong software engineering fundamentals with experience building production applications.
• Hands-on experience developing LLM, Generative AI or agent-based systems.
• Experience building AI-powered applications that extend beyond basic model or API integrations.
• Practical experience designing prompts, agent workflows, evaluations and AI behaviour.
• Understanding of tool calling, context engineering, retrieval, memory and multi-step agent execution.
• Ability to write clean, maintainable and production-quality code.
• Comfortable working across multiple abstraction layers, from model → system → product.
• Strong understanding of API architecture and backend systems.
• Strong problem-solving skills and the ability to operate effectively within ambiguous technical environments.
• A bias toward experimentation, shipping, iteration and continuous improvement.

Technology Environment

• Python
• Large Language Models (LLMs)
• Commercial and open-weight models
• LLM APIs and model providers
• Agent frameworks and orchestration systems
• Vector databases
• Retrieval systems / RAG
• AI memory and context systems
• Tool calling and structured outputs
• Backend services and APIs
• Distributed systems
• PyTorch
• JAX
• AI evaluation and observability tooling

What Success Looks Like

You will successfully take AI capabilities from experimentation into reliable production experiences that deliver measurable user value. LLM-powered workflows will become increasingly scalable, observable and maintainable, with systematic evaluation and experimentation used to improve AI quality over time. Agent workflows will become more predictable, efficient and cost-effective while maintaining strong performance across complex, multi-step tasks. Ultimately, you will help translate sophisticated AI capabilities into simple and intuitive product experiences, ensuring the complexity behind the technology is largely invisible to the end user.