Skip to main content
B

BabylonAI

BabylonAI provides a logging platform for computer vision teams that captures raw input and prediction data from edge devices, enabling efficient data organization by model version and device. This solution addresses the challenge of evaluating model performance in production environments, particularly in low-connectivity situations, while facilitating the collection of relevant data for retraining.

Dover, United KingdomFounded 202321K+ followers
Updated 20 months ago

Funding

$800K raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.

EFKVOPPI
Funding rounds are not available yet.

Founders

Product

Problem

Computer vision teams face challenges in evaluating model performance in production environments, especially on edge devices with limited connectivity. Collecting and organizing relevant data for model retraining from these deployments is often inefficient and difficult.

Solution

BabylonAI offers a telemetry platform designed for computer vision teams to efficiently manage the post-deployment data flow required for model evaluation and retraining. The platform captures raw input and prediction data directly from edge devices, organizing it by model version, device, and timestamp. This enables teams to quickly assess model performance in production, identify outlier samples, and trigger the collection of specific data subsets for retraining. BabylonAI supports deployments in low-connectivity environments and integrates with existing infrastructure, such as AWS or local servers, using high-quality TCP streaming.

Target Audience

The primary users are computer vision teams deploying models on edge devices, including those in on-prem server, phone, and other edge device deployments.

Features

  • Logging of raw input and prediction data from edge devices
  • Data organization by model version, device, and timestamp
  • Filtering of production data on-device
  • Support for Python, C++, Swift, Flutter, and React Native
  • Integration with AWS and local servers
  • High-quality streaming via TCP
  • Custom smart triggers for targeted data collection
  • Open-source logger, babylog
This profile is AI-generated and may contain inaccuracies.