Astera Holdings

Compensation

Full Description

Head of Computer Vision Location: Remote (NYC Preference) Type: Full-Time Reports to: CEO About Astera Holdings Astera Holdings is building the intelligence architecture for prediction itself — a quantitative infrastructure designed to transform how humanity perceives, prices, and acts on uncertainty. Our ecosystem operates through three synergistic pillars, each reinforcing a unified mission to institutionalize foresight as an investable science: • Atlas (B2C) — democratizing event investing and transforming speculation into structured, data-driven decision-making for the global market. • Astera Capital (Hedge fund) — establishing prediction markets as a recognized, institutional asset class through disciplined quantitative strategies. • Astera Analytics (Enterprise AI B2B) — extending our predictive intelligence to solve complex, real-world problems across industries. At the core of this ecosystem is the Reflexivity Engine — Astera’s continuously learning system that captures, models, and predicts real-world outcomes across sports, economics, and geopolitics. Astera’s mission extends beyond forecasting. We are engineering the underlying science of foresight itself — the architecture through which the future becomes knowable, actionable, and investable. Role Overview Astera is seeking a Head of Computer Vision to lead the company’s efforts to extract qualitative intelligence from the world’s most information-dense medium — video. This is not a role confined to object detection or tracking. It is an opportunity to teach machines to see meaning — to interpret context, intent, and probability from raw pixels. Your mandate is to design systems that transform unstructured visual reality into structured, predictive intelligence capable of understanding human behavior, strategy, and decision-making in real time. You will architect and lead the development of Astera’s visual intelligence infrastructure, building models and pipelines that become the perceptual foundation of our Reflexivity Engine. Your work will bridge computer vision, reinforcement learning, and causal inference, advancing the frontier where perception and prediction converge. Key Responsibilities Vision Intelligence Strategy: • Define and execute Astera’s computer vision strategy — from research priorities to system architecture — ensuring alignment with the company’s broader mission of predictive intelligence. Evaluate build-versus-partner decisions and establish clear technical milestones for model deployment and feature integration. System Architecture & Implementation: • Design and implement the end-to-end vision pipeline, capable of ingesting live video streams, processing frames at scale, and outputting structured, real-time data optimized for latency, accuracy, and reliability. Model Research & Development: • Develop, train, and fine-tune state-of-the-art computer vision models — including convolutional and transformer-based architectures — for tasks such as multi-object tracking, pose estimation, and event segmentation. Ensure that models are robust, generalizable, and production-ready. Cross-Domain Integration: • Collaborate closely with AI, quant, and data engineering teams to integrate visual intelligence into the Reflexivity Engine — enriching multimodal models that predict real-world outcomes across sports, finance, and geopolitics. Leadership & Team Formation: • Serve as Astera’s internal authority on computer vision. Build and lead a small, high-caliber team of CV researchers and engineers, fostering a culture of scientific rigor, speed, and intellectual honesty. Continuous Innovation: • Stay at the frontier of computer vision and multimodal AI research. Proactively identify and adapt emerging methods — from self-supervised learning to causal perception — that can expand Astera’s predictive edge. Qualifications • Required: experience in computer vision or related AI fields, with a track record of building or deploying CV systems in production. • Required: Deep knowledge of image processing and deep learning for vision (CNNs, object detection like YOLO/RCNN, pose estimation, etc.). • Required: Proficiency in Python, C++, and common CV/ML libraries (OpenCV, TensorFlow/PyTorch). • Required: Experience architecting systems that process data in real-time or near-real-time where latency is critical. • Required: A strong grasp of the underlying mathematics of computer vision and machine learning. • Required: A demonstrated ability to solve open-ended problems with creativity and persistence. • Preferred: Prior experience in sports analytics or familiarity with sports data. • Preferred: A Master's or PhD in Computer Vision, Machine Learning, or a related field. • Preferred: Experience with deploying CV solutions on cloud platforms (using GPU instances, video streaming services, etc.). Required: GRIT and OBSESSION To Apply careers@astera.holdings