Snorkel AI: Blowing out labeling bottlenecks for machine learning data

Snorkel AI logo Intellyx BCAn Intellyx Brain Candy Brief

Snorkel AI focuses on eliminating the constraints of labeling a flow of unstructured and structured training data for use in machine learning and AI recognition scenarios.

The hardest part of ML training models involves reaching a high enough degree of confidence in how an incoming data fragment or document will be recognized and classified with a label by AI engines. Usually, a high degree of ‘hand labeling’ and manual maintenance of existing labels is required on the part of human subject matter experts, as the validity of the initial training data degrades over time.

The fast-growing academically-founded startup’s Snorkel Flow platform incorporates an open source labeling core under a massively parallel neural network ‘coach’ that combines and compares many different labeling functions, using incoming data to code each asset with an ontology in a process they call ‘data programming.’

A slick monitoring dashboard allows the owners of enterprise AI training data to analyze the performance of labeling functions by measuring data confidence and success ratings.

©2021 Intellyx, LLC. At the time of writing, Snorkel AI is not an Intellyx customer. Want to see more BrainCandy? Subscribe today. Get our Cloud-Native Computing poster. If you are a vendor seeking coverage from Intellyx, please contact us at PR@intellyx.com.

SHARE THIS:

Principal Analyst & CMO, Intellyx. Twitter: @bluefug