I am Diego Pousa, a specialized data collection vendor focused on ego-centric video data sets for computer vision projects. I lead a pre-vetted team of contributors and coordinate data collection across diverse environments to deliver high-quality, high-volume datasets. I have a proven track record establishing internal QA pipelines to achieve 100% acceptance rates for AI training data. I implement secure, efficient workflows such as a single point of entry to protect accounts and data integrity, and I manage hardware logs, logistics, and payout structures for the local team to keep operations smooth and compliant.…

Diego Pousa

I am Diego Pousa, a specialized data collection vendor focused on ego-centric video data sets for computer vision projects. I lead a pre-vetted team of contributors and coordinate data collection across diverse environments to deliver high-quality, high-volume datasets. I have a proven track record establishing internal QA pipelines to achieve 100% acceptance rates for AI training data. I implement secure, efficient workflows such as a single point of entry to protect accounts and data integrity, and I manage hardware logs, logistics, and payout structures for the local team to keep operations smooth and compliant.…

Available to hire

I am Diego Pousa, a specialized data collection vendor focused on ego-centric video data sets for computer vision projects. I lead a pre-vetted team of contributors and coordinate data collection across diverse environments to deliver high-quality, high-volume datasets.

I have a proven track record establishing internal QA pipelines to achieve 100% acceptance rates for AI training data. I implement secure, efficient workflows such as a single point of entry to protect accounts and data integrity, and I manage hardware logs, logistics, and payout structures for the local team to keep operations smooth and compliant.

See more

Work Experience

Lead Data Collection Vendor at Confidential AI Project
January 1, 2024 - Present
Managed a team of contributors collecting ego-centric video data for a major US-based AI Robotics project. Implemented a strict single point of entry system to maintain account security and data integrity. Achieved high-quality metrics by filtering out shaky, low-light, or irrelevant footage prior to submission. Managed hardware logs, logistics and payout structures for the local team.

Education

Add your educational history here.

Qualifications

Add your qualifications or awards here.

Industry Experience

Media & Entertainment, Software & Internet, Professional Services
    uniE608 High-Fidelity POV Sample: Outdoor Patio Maintenance & HDR Lighting
    Dataset Sample: Raw ego-centric (POV) footage capturing maintenance tasks in an outdoor patio environment. Technical Complexity: High Contrast Lighting (HDR): Demonstrates sensor performance in direct sunlight with hard shadows (critical for outdoor computer vision). Fluid Dynamics: Interaction with water, foam, and cleaning tools in an uncontrolled environment. Environmental Elements: Exposure to wind and outdoor acoustics. Use Case: Ideal training data for Outdoor Service Robots, Environmental Perception Models, and autonomous cleaning systems. Production Capacity: My team operates in diverse semi-outdoor and outdoor residential perimeters (Gardens, Patios, Driveways). Outdoor NaturalLight HDR Robotics Cleaning video…
    uniE608 High-Fidelity POV Sample: Delicate Object Handling & Sorting
    Dataset Sample: Raw ego-centric (POV) footage capturing the disassembly and sorting of holiday decorations. Technical Complexity: Fragile Object Handling: Demonstrates careful manipulation of glass/delicate ornaments (Grip force estimation data). Deformable Linear Objects (DLOs): Untangling and winding light strings/wires—a critical benchmark task for robotic manipulation. Occlusion: High levels of object overlap (branches vs. ornaments) providing robust training data for segmentation models. Use Case: Highly relevant for Logistics AI, Warehouse Packing Robots, and Fine Motor Control training. Production Capacity: My team can replicate complex sorting and packing tasks across varying object types and environments. Robotics Logistics FineMotorSkills Sorting ComputerVision video…
    uniE608 High-Fidelity POV Sample: Household Maintenance & Reflective Surfaces
    Dataset Sample: Raw ego-centric (POV) footage capturing household maintenance tasks in a bathroom environment. Technical Complexity: Reflective Surfaces: Demonstrates clear data capture handling specular reflections (mirrors, chrome faucets, ceramic) which are critical edge-cases for Computer Vision. Deformable Object Interaction: Manipulation of cleaning cloths and fluids. Fine Motor Control: Scrubbing, wiping, and rinsing actions. Use Case: Essential training data for Service Robotics, Home Automation AI, and Activity Recognition models. Production Capacity: My team can scale this specific "Home Service" dataset to hundreds of hours, covering various bathroom layouts and cleaning protocols. Robotics HomeService Cleaning ComputerVision EdgeCaseData video…
    uniE608 High-Fidelity POV Sample: Complex Hand-Object Interaction (Kitchen)
    Dataset Sample: Raw ego-centric (POV) footage demonstrating fine motor skills and appliance interaction in a kitchen environment. Technical Specs: Resolution: High Definition (Raw export) Perspective: First-Person (Head-mounted stabilization) Scenario: Multi-step food preparation (Electric Juicer usage) Why this matters for AI: This footage is optimized for training Computer Vision models in Human-Object Interaction (HOI), Action Recognition, and Robotic Imitation Learning. It features clear occlusion handling (hands over objects) and stable framing essential for annotation. Production Capacity: I manage a team equipped to generate 50+ hours per week of similar task-based scenarios across diverse environments (Kitchen, Garage, Retail, Outdoor). ComputerVision video DataCollection POV MachineLearning VideoProduction…