AI Engineer – R&D

📋 Description

• Design and implement strategies for fine-tuning structured generation and editing workflows.

• Create supervised datasets derived from successful generations, retries, failures, and user modifications.

• Establish measurable benchmarks to assess generation quality, correctness, and the preservation of edits.

• Conduct experiments with open-source models like Llama, Qwen, Mistral, DeepSeek, or similar architectures.

• Apply techniques such as LoRA, QLoRA, supervised fine-tuning (SFT), distillation, preference tuning, or synthetic data methods as suitable.

• Develop automated pipelines for gathering, cleansing, evaluating, and incorporating production data into training datasets.

• Utilize validation systems, intermediate representations, runtime analysis, and rendered outputs as structured feedback for models.

• Enhance retry, repair, and self-correction workflows within generation pipelines.

• Work collaboratively with engineering and product teams to boost model reliability and output quality.

⛳️ Requirements

• Proven experience in building with LLMs or structured generation systems in production or applied research environments.

• Practical experience in fine-tuning or adapting open-source language models.

• Proficient Python programming skills.

• Background in developing evaluation systems, ML experimentation workflows, or data pipelines.

• Solid understanding of prompt engineering, structured outputs, tool utilization, and model failure analysis.

• Capability to define measurable evaluation criteria instead of depending solely on subjective assessments.

• Proficient in troubleshooting systems that encompass model outputs, validation systems, runtime behavior, and rendered results.

• Excellent communication and teamwork abilities.

• Experience with code generation, DSL generation, or AI systems that are aware of compilers.

• Familiarity with LoRA, QLoRA, SFT, preference tuning, distillation, or synthetic data generation techniques.

• Knowledge of animation systems, graphics pipelines, design tools, SVG, WebGL, shaders, or procedural graphics.

• Experience with multimodal or visual-language-model evaluation workflows.

• Familiarity with observability or ML evaluation tools like Weights & Biases, Langfuse, MLflow, or OpenTelemetry.

• Experience in building agentic systems, orchestration pipelines, or multi-step generation workflows.

• Understanding of ASTs, intermediate representations (IRs), or structured program representations.

🏝️ Benefits

• Opportunity to work on innovative projects in a cutting-edge field.

• Collaborative and inclusive work environment.

• Professional development and learning opportunities.

• Competitive salary and comprehensive benefits package.