Senior Computer Vision Engineer

Arrive AI Inc•Fishers, IN
•Onsite

About The Position

Arrive AI is seeking a Senior Computer Vision Engineer to own the Arrive Point vision pipeline end to end, from camera input and edge-device inference to events consumed by ArriveOS and data sent to the cloud. The role selects, trains, fine-tunes, and optimizes models, deploys them to NVIDIA Jetson hardware, and maintains their accuracy in the field. It is a hands-on senior role focused on edge devices, with work in C++ and Python using NVIDIA DeepStream, GStreamer, Triton Inference Server, TensorRT, and related frameworks. The engineer will collaborate with the platform team on operating-system and cloud services. The position is based on-site in Fishers, Indiana, or nearby. The engineer will join Arrive AI’s Product Team, a fast-paced team developing Arrive Point units, networks, and AI-powered platforms for autonomous last-mile logistics. The role may require travel up to five days per month. Arrive AI states that it cannot provide visa sponsorship for this position. Compensation is described as market competitive with a high equity incentive.

Requirements

  • Master’s degree in Computer Science, Computer Vision, Machine Learning, Electrical Engineering, or a related field and 3–5 years of industry experience, or a PhD and at least 3 years of industry experience.
  • Hands-on experience with real cameras and edge hardware is heavily valued.
  • Track record of shipping a vision system from camera through production deployment, rather than only training models.
  • Experience with classical and modern computer vision, including evaluating, selecting, training, and fine-tuning classifiers, CNNs, vision transformers, object detectors, and vision-language models in PyTorch.
  • Experience with TensorRT and ONNX, including engine builds, quantization, and accuracy checks after conversion.
  • Experience with NVIDIA DeepStream and GStreamer.
  • Experience with Triton Inference Server, gRPC, and modern API-driven development workflows.
  • Experience with NVIDIA Jetson hardware, including Orin and Thor, and other edge compute such as Raspberry Pi.
  • Strong C++ skills (C++17 or later) and Python skills.
  • Ability to collect data on edge devices, including sampling, curation, labeling, edge-case analysis, and adjusting camera and lighting conditions.
  • Experience with Docker, Kubernetes or K3s, CI/CD pipelines, and working familiarity with AWS or Azure and S3-compatible object storage.
  • Clear writing and presentation skills for technical and non-technical audiences, with a habit of documenting work.
  • Ability to learn new tools, platforms, and problems quickly in a fast-moving startup.

Nice To Haves

  • Experience with YOLO, TinyViT, NVIDIA PeopleNet, or the NVIDIA TAO Toolkit.
  • Experience with WebRTC live video, including SFUs, WHIP, and TURN and STUN.
  • Familiarity with Unix domain sockets, shared memory, and message framing with FlatBuffers or Protocol Buffers.
  • Knowledge of image sensors, optics, housings, GigE, MIPI CSI-2, USB, V4L2, Intel RealSense, or time-of-flight sensors.
  • Familiarity with Rust-based architectures or Go.
  • Experience with OCR or depth sensing.
  • Experience with dataset versioning, drift monitoring, or label quality analysis.
  • Experience with RTSP, ONVIF, or video management systems.
  • Background in logistics, delivery, retail, or security video.
  • HIPAA familiarity or experience with on-premises-only systems.

Responsibilities

  • Own the Arrive Point vision stack from camera ingress through inference to events consumed by ArriveOS and data reaching the cloud.
  • Develop and maintain C++ pipelines using NVIDIA DeepStream and GStreamer on Jetson Orin and Thor, with Triton Inference Server and TensorRT.
  • Evaluate, select, train, and fine-tune models for different tasks, including classifiers, CNNs, vision transformers, object detectors, and vision-language models.
  • Export models through ONNX to TensorRT, build per-platform FP16 and INT8 engines, and verify device accuracy, latency, and memory use.
  • Sample, curate, and label data from deployed cameras, refine camera configurations, and address edge cases.
  • Maintain the model training and release lifecycle, including dataset versioning, CI-gated evaluation, ONNX and TensorRT parity checks, robustness testing, drift monitoring, and retraining.
  • Work with the platform team on gRPC contracts between the vision service and ArriveOS, including event debouncing, health reporting, and inter-process communication.
  • Maintain live video egress over WebRTC for remote viewing and operations.
  • Containerize and deploy vision services to edge devices with Docker and K3s, including cross-builds, hardware-in-the-loop tests, and simulated test environments.
  • Collaborate with cloud and platform team members on object storage, model publishing, and data flows between edge devices and the cloud.
  • Design camera-data handling with privacy and security in mind, including consideration of video containing personal information.
  • Write architecture decision records, technical guides, and test reports, and present results and trade-offs to engineers, leadership, and partners.
  • Perform other duties as assigned.

Benefits

  • High equity incentive; the company states that every employee is an equity holder and a business owner.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service