[Remote] AI Research Engineer

Note: The job is a remote job and is open to candidates in USA. DeepHow is a Physical AI platform for industrial manufacturing, pharmaceuticals, and energy, focused on capturing expert know-how and driving verified execution. They are seeking an AI Research Engineer to innovate across their AI platform by researching and developing next-generation models and techniques, working closely with product, engineering, and AI teams.ResponsibilitiesResearch and development of AI models for industrial knowledge capture, verification, and workflow intelligenceExperimentation with LLMs, vision-language models (VLMs), speech models, and multimodal systems — including evaluation of open-weight and proprietary models for customer-specific constraintsDesign and evaluation of advanced RAG architectures, agentic workflows, and reasoning systemsModel training, fine-tuning, and optimization pipelinesBenchmarking, evaluation, and experimentation frameworks that measure both model quality and business impactAssessment of emerging AI techniques and their applicability to DeepHow’s platform and roadmapCollaboration with engineering to transition research prototypes into scalable production systemsInternal research write-ups, technical documentation, and best practicesSkillsBachelor's, Master's, or Ph.D. in Computer Science, Machine Learning, AI, or a related technical field — or equivalent practical experience3+ years of experience in AI/ML research, applied AI, or machine learning engineeringStrong Python skills with deep expertise in PyTorch (or TensorFlow) and the Hugging Face / Transformers ecosystemHands-on experience training, fine-tuning, and evaluating foundation modelsPractical experience with LLMs and VLMs, prompt engineering, RAG systems, and rigorous model evaluation methodologiesFamiliarity with vector databases and embedding systemsExperience with large datasets and distributed training and inference environmentsStrong grasp of ML fundamentals, deep learning, and modern AI architectures, plus solid experiment design and statistical analysisA track record of translating research concepts into shipped product featuresExperience with video understanding, computer vision, speech recognition, or other multimodal AIFamiliarity with reinforcement learning, agentic systems, or reasoning modelsExperience with MLflow, Kubeflow, Ray, Airflow, or similar ML platformsPublications, patents, open-source contributions, or active participation in AI research communitiesExperience deploying AI systems on GCP or AWSBackground in manufacturing, industrial automation, workforce training, or operational technologyExperience building AI products in an enterprise SaaS environmentBenefitsFull-timeRemote (Dallas preferred)Access to cutting-edge AI technologies, hardware, and research opportunitiesHigh ownership, rapid experimentation, and fast decision-makingA talented, mission-driven team applying AI to problems that matterInfluence over the AI roadmap, from research through productionCompany OverviewDeepHow develops an AI-powered learning platform for manufacturing, service, and repair. It was founded in 2018, and is headquartered in Royal Oak, Michigan, USA, with a workforce of 51-200 employees. Its website is https://www.deephow.com.