Large Language Model Training

Foundation and Medium LLM pretraining, finetuning, and contrastive learning pipelines

Our team has experience in all domains of the LLM stack, from dataset filtration and deduplication to creating pretraining and finetuning recipes for maximal task effectiveness

Large Language Model Deployment

We deploy Large Language Models for customers in their environments using SOTA practices

Our team has experience in using and developing SOTA techniques for latency reduction, model compression, and model deployment on ROCm, CUDA, Tranium, TPUs, and more

Manufacturing and Robotics

We build Computer Vision systems for Manufacturing

We love tinkering and interacting with the physical world through Computer Vision and Robotics. From tracking to localization and segmentation, we've built systems for a variety of companies focusing on real-world problems

Defense

We work with Autonomous Systems and Agents

Our team has direct experience in autonomous systems and agents for defense applications using SLAM, path planning, and coordination