About this opportunity
Overview
Appen is hiring an Applied AI Research Engineer for its AI Data Opportunities team, focused on applied research and implementation work that supports frontier lab initiatives and customer engagements. This role is hands-on and suited to someone who enjoys turning research ideas into practical systems that can be reused across multiple opportunities.
The position is centered on building, testing, documenting, and operationalizing AI research assets in collaboration with the GenAI Research team and cross-functional stakeholders.
What You'll Do
- Build reinforcement learning and agent environments for customer and frontier lab use cases, including task specifications, scoring, and evaluation.
- Develop benchmarks and evaluation harnesses to measure model and data quality across dimensions such as accuracy, robustness, safety, latency, and cost.
- Create LLM pipelines and agentic systems that support research, evaluation, and customer trials.
- Run fine-tuning, adapter, and related model experiments to understand how data and methods affect model behavior.
- Deploy local or self-hosted models for evaluation, inference, and automation workflows.
- Document experiments, configurations, data, results, and known limitations so others can reproduce and extend the work.
- Partner with research and cross-functional teams to turn technical work into reusable assets for customer engagements.
Requirements
- Bachelor’s, Master’s, or PhD in Computer Science, Engineering, Machine Learning, or a related technical field.
- At least 3 years of professional engineering or relevant industry experience in AI/ML or software engineering.
- Strong software engineering skills with experience building reliable and maintainable AI systems.
- Hands-on experience with agentic systems, reinforcement learning environments, LLM pipelines, or similar AI systems.
- Experience building evaluation harnesses, benchmarks, or model testing pipelines.
- Ability to work independently on technical problems and move quickly from an idea or research question to a working solution.
- Strong understanding of experimentation, reproducibility, and technical documentation.
Nice to Have
- Experience developing synthetic data generation systems or datasets.
- Published research papers, benchmarks, or other technical research.
- Experience with SWE-bench or similar software engineering evaluation environments.
- Experience building or deploying local inference, open-weight models, or self-hosted model environments.