• Fully remote position

• Candidates should have at least 4 hours of overlap with the Malaysia time zone (MYT / UTC+8)

• Direct hire opportunity

• Visa sponsorship is not available for this role

About the Company

We are partnering with a fast-growing technology company building next-generation AI-powered creative tools for animation, design, and interactive media workflows. Their platform is used to help creators and teams accelerate visual content production through intelligent generation, editing, and automation systems.

The company operates at the intersection of generative AI, structured content systems, and creative tooling, with a strong focus on production-grade AI infrastructure, model reliability, and real-world usability. Their engineering culture emphasizes experimentation, measurable quality improvements, and scalable AI systems deployed directly into user-facing products.

About the Role

We’re looking for an AI Engineer to help build specialized generation, editing, evaluation, and optimization systems for creative and structured content workflows.

This role focuses on structured generation, domain-specific model adaptation, evaluation systems, feedback pipelines, and production AI infrastructure. You’ll work closely with engineering, design, and product teams to improve generation quality, reliability, efficiency, and usability across AI-assisted creative workflows.

This is an opportunity to work on highly applied AI problems involving:

• Structured and constrained generation

• AI-assisted editing systems

• Evaluation and observability pipelines

• Fine-tuning and adaptation of open-source models

• Multi-step generation and repair workflows

• Scalable production AI systems

What You’ll Work On

• Natural-language-to-structured-content generation workflows

• Structure-preserving editing and modification systems

• Validation and repair pipelines for generated outputs

• Evaluation systems for quality, correctness, consistency, and runtime performance

• Training and evaluation datasets built from production usage and interaction traces

• Smaller, lower-latency models for targeted generation, editing, routing, and repair tasks

• Multi-step orchestration and self-correction workflows for AI systems

Key Responsibilities

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

• Build supervised datasets from successful generations, retries, failures, and user edits

• Develop measurable benchmarks for generation quality, correctness, and edit preservation

• Experiment with open-source models such as Llama, Qwen, Mistral, DeepSeek, or related architectures

• Implement LoRA, QLoRA, supervised fine-tuning (SFT), distillation, preference tuning, or synthetic data approaches where appropriate

• Build automated pipelines for collecting, cleaning, evaluating, and promoting production data into training datasets

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

• Improve retry, repair, and self-correction workflows for generation pipelines

• Collaborate cross-functionally with engineering, design, and product teams to improve model reliability and output quality

Required Qualifications

• Strong experience building with LLMs or structured generation systems in production or applied research settings

• Hands-on experience fine-tuning or adapting open-source language models

• Strong Python engineering skills

• Experience building evaluation systems, ML experimentation workflows, or data pipelines

• Strong understanding of prompt engineering, structured outputs, tool use, and model failure analysis

• Ability to define measurable evaluation criteria rather than relying only on subjective review

• Comfort debugging systems spanning model outputs, validation systems, runtime behavior, and rendered results

• Strong communication and collaboration skills

Preferred Qualifications

• Experience with code generation, DSL generation, or compiler-aware AI systems

• Experience with LoRA, QLoRA, SFT, preference tuning, distillation, or synthetic data generation

• Familiarity with animation systems, graphics pipelines, design tools, SVG, WebGL, shaders, or procedural graphics

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

• Experience with observability or ML evaluation tooling such as Weights & Biases, Langfuse, MLflow, or OpenTelemetry

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

• Familiarity with ASTs, intermediate representations (IRs), or structured program representations

Why This Role Is Interesting

• Work on real-world AI systems used in creative production workflows

• Build beyond prompt engineering into evaluation, repair, optimization, and infrastructure

• Help shape production-grade AI systems for next-generation creative tooling

• Collaborate with a highly technical and product-focused engineering team

• Tackle challenging problems involving structured generation, multimodal systems, and AI reliability