Hire this AI Trainer
Sign in or create an account to invite AI Trainers to your job.
Vision intelligence platform (Payana) with human-in-the-loop data labeling for multiple modalities including text
Bengaluru, India
CVAT
Labelbox
Surge AI
Roboflow
Data Annotation Tech
Mercor
LabelImg
OneForma
Image
Video
Geospatial Tiled Imagery
Text
Polygon
Bounding Box
Segmentation
Classification
Object Detection
Data CollectionWe're a software factory that builds solutions for businesses around the world. We work on some of the deep tech problems using our expertise in data annotation, AI/ML training models, vision intelligence, application modernization, digital process automation, digital supply chain among other.
Encryption Everything: Utilizing AES-256 encryption for data at rest and TLS 1.3 for data in transit. No Local Storage (Cloud-Only): Using a secured centralized data labeling platform (e.g., CVAT, OneTrust integrated workflows) that blocks annotators from downloading, capturing screenshots, or copying data to personal devices. 4. Workforce & Physical Facility Security How the workforce is managed introduces the greatest vulnerability (malicious intent or human error).In-house/Managed Teams vs. Crowdsourcing: For highly sensitive datasets, the industry gold standard is a fully vetted, in-house, or in-facility workforce bound by strict Non-Disclosure Agreements (NDAs). Crowdsourcing presents significantly higher risk.Clean Room Facilities: For maximum security (like defense, automotive TISAX, or proprietary IP), teams operate in physical clean-room offices where personal electronics, smartphones, and USB drives are prohibited, and monitors use polarized privacy filters
CVATVideoPolygonClassificationData labeling and annotation for car driving through traffic, the goal is to keep annotating different objects and variables while driving. This project is for a self driving use case for an automobile company.
CVATVideoBounding BoxPolygonThis project was to annotate the electric switch detection model and train it for automating defect detection in real time through camera.
TextData CollectionThe resume describes building a vision intelligence platform that uses human-in-the-loop labeling to train accurate AI models on a wide range of data types, including text. It focuses on leveraging expert human loop data labeling to improve model training quality and downstream performance. The described work is positioned as an AI training capability within the product offering. • Human-in-the-loop data labeling approach. • Training data includes text (and other modalities such as images, audio, video, and LiDAR). • Aim is to produce accurate AI models through curated labeled data. • AI training via expert labeling rather than automated-only pipelines.