Artificial Intelligence Engineer

Principal AI/ML Engineer - Build the model, not just the wrapper Austin, TX | Senior to Principal level | Full-Time | Hybrid Working This is for the engineer who wants to work closer to the model. I’m supporting an Austin-based AI company building domain-specialized models and autonomous systems for complex, regulated operations.

The technology is already processing real decisions in production, where accuracy, latency, explainability and auditability all matter. The next hire will help determine how the underlying models are trained, evaluated, optimized, served and continuously improved. The engineering questions you’ll own How do we turn large volumes of expert-reviewed decisions into high-quality training and evaluation data?

When should we fine-tune, distill, route or replace a model? How do we improve reasoning without sacrificing latency or predictability? How do we evaluate performance beyond a generic benchmark?

How do we identify failure patterns and turn them into better training data? How do we increase throughput and GPU utilization while reducing serving cost? How do we build autonomous workflows that remain controlled, explainable and auditable?

What you could be building - Depending on your strengths, the work may include: Training and fine-tuning foundation and smaller language models Post-training through SFT, preference optimization or reinforcement-learning approaches Domain-specific dataset creation, labeling and synthetic-data generation Model evaluation, benchmarking, red-teaming and failure analysis Distillation, quantization and inference optimization High-performance model serving using technologies such as vLLM, Triton or TensorRT-LLM Distributed training and inference across GPU environments Retrieval and agentic systems with validation, guardrails and human review Production monitoring across quality, drift, latency, throughput and cost Converting real production failures into continuous model improvements The environment includes Python, PyTorch, Hugging Face, CUDA, distributed GPU infrastructure, modern inference runtimes, cloud platforms and production observability .

What matters more than your job title

You may currently be an AI Engineer, Machine Learning Engineer, Applied Scientist, Research Engineer, LLM Engineer or ML Infrastructure Engineer.

The title is less important than evidence that you have personally: Trained, fine-tuned or materially optimized models Built evaluation systems and investigated why models fail Taken AI beyond experimentation and into production Worked through performance, scaling or inference constraints Made technical decisions involving data quality, model quality and system reliability Owned meaningful engineering outcomes rather than only contributing around the edges There is plenty of application-layer engineering involved, but the center of gravity is the models, data, evaluation and production infrastructure underneath them.

Why consider it?

Principal-level influence without moving away from hands-on engineering Greenfield development alongside established production systems Direct access to experienced, technically involved AI leadership Complex proprietary data and real-world model feedback Meaningful ownership over architecture and technical direction The opportunity to shape a growing AI engineering function in Austin Work where improvements to quality, speed and reliability have measurable consequences