Forward Deployed AI Engineer

Lunon; Oregon; Hybrid; USD 180,000 - 240,000; Full time

Job summary

Lubon? Lunon is seeking a Forward Deployed AI Engineer to define how intelligence enters a company’s product. You’ll choose models, tailor them for ERP contexts, and balance cost and latency to deliver AI-native capabilities.

You’ll join a small, craft-focused team backed by strong investors, shipping AI features that influence real client outcomes and workflows across production systems.

Qualifications

• Experience shipping models or features in production.
• Ability to explain what worked and what didn’t in fine-tuning or post-training.
• Experience designing evaluations and benchmarks that drive decision-making.

Responsibilities

• Choose the right model for each problem, from frontier APIs to open-weight models, and know when fine-tuning outperforms a better prompt
• Fine-tune where it pays off: supervised fine-tuning, preference optimization, and parameter-efficient methods such as LoRA
• Design the evaluations and benchmarks that show whether a change made things better
• Take models from training to serving, balancing cost, latency, and quality
• Work with the company's team to find where AI will change their results most, and decline work that won't repay the effort

Skills

LLM fine-tuning

Model evaluation

Production deployment

Cost & latency optimization

AI tooling familiarity

Forward deployment

Tools

AI coding tools

Job description

ABOUT LUNON

The companies that lead the next decade won't adopt AI, they'll be rebuilt around it. Lunon treats strategy and software as one discipline, shaping the systems that let ambitious companies outthink their markets and outpace their competitors.Most advice about AI ends as a recommendation that someone else has to carry out. At Lunon, the people who decide what should change are the people who build it, so an insight becomes capability rather than a slide. We are backed by General Catalyst, and our open-source research system ranks first on DeepResearch Bench.We're a small team and intend to stay that way. Everyone here owns their work from the first conversation to production, adopts the newest models as they land, and holds an exacting standard for craft. If you want your work to change how a company operates, not just inform it, we'd like to meet you.

ABOUT THE ROLE

As a Forward Deployed AI Engineer, you will decide how intelligence enters a company's product: which model, adapted how, proven by what evidence, and at what cost.You will work alongside a Forward Deployed Data Engineer, who owns the data foundation. Your first project: helping an ERP company make its product AI-native.

RESPONSIBILITIES

• Choose the right model for each problem, from frontier APIs to open-weight models, and know when fine-tuning outperforms a better prompt
• Fine-tune where it pays off: supervised fine-tuning, preference optimization, and parameter-efficient methods such as LoRA
• Design the evaluations and benchmarks that show whether a change made things better
• Take models from training to serving, balancing cost, latency, and quality
• Work with the company's team to find where AI will change their results most, and decline work that won't repay the effort

YOU MAY BE A FIT IF

We weigh what you have shipped and sustained over years of experience.

• You have fine-tuned or post-trained a model and can explain what worked and what didn't
• You have built evaluations that changed a decision
• You have shipped LLM features that people depend on
• You have served a model in production and owned its cost and latency
• You work with AI coding tools every day and judge the result by the system it produces

STRONG CANDIDATES MAY ALSO

• Have built on or integrated with an ERP, CRM, or accounting system
• Have shipped extraction or search over complex business documents
• Have done consulting or forward-deployed work, and know how to earn trust inside a company that isn't yours

LOCATION

Remote within the US. Lunon is headquartered in New York City.

HOW WE HIRE

• A 30-minute conversation with the team about what you've built and what we're building.
• A technical deep-dive into one system you owned, from the business problem to production.
• A working session with the team on a real client problem.
• References, then an offer. The process takes about two weeks.