top of page

AI that Understands and Explains

Easy to use platform for predictive insights and simulation 

We'll be in touch soon.

Why Metonymize?

The Metonymize platform enables the combination of 'thinner' data with domain knowledge to produce superior explainable AI

EXPLAINABLE

Generative modelling approach to AI  offers ground-up support for auditability (Anti "black box")

KNOWLEDGE

Simulations capture human expertise 

(Intuition, Judgement)

THIN DATA

Next generation AI compensates for limited data - generalizing insights via Transfer Learning

Solutions: How enterprises use Metonymize

Traditional Machine Learning only works with lots of data. In reality, predictive systems can combine limited data with expert knowledge using advanced AI techniques

NEW PRODUCT DEV.

Transfer insights for better planning​​

Consumer behavior models for concept testing at scale using simulation. Scale up experimentation while reducing marginal cost to zero.

SURVEY ANALYTICS

User-driven insights from limited data 

Extend limited datasets using domain knowledge. Analyze hypothetical scenarios and conduct what-if analysis.

VOICE OF CUSTOMER

Consumer experience optimization

Automatically process natural language feedback in real-time. Drive operational response times down to zero to improve customer NPS.

Solutions

Metonymize Platform

Full-stack architecture, easily integrated through APIs and enterprise standard web/mobile Apps

Support industry standard ETL pipelines and REST API based data input/output

DATA IO
NATURAL LANGUAGE ENGINE

5X better accuracy in language understanding.No “Cold Start” problems given pre-trained language models of semantics from different domains

DOMAIN KNOWLEDGE STACK

Probabilistic programing libraries and tooling make it possible to seamlessly inject expert knowledge

lack of granular data to train machine learning models

SIMULATOR FACTORY
& WAREHOUSE

​Zero data generative models capture expert domain knowledge as a starting point for machine learning

Third Wave AI (3Wai) architecture supports the injection of domain knowledge to tune models with expert context
 

3Wai algorithms use computational capacity to “trade off” model accuracy and search time with/without more data

THIRD WAVE AI
MICROSERVICES

Scalable cloud-based High-Performance GPU enabled Compute stack for model training and efficiency


Probabilistic Computing stack delivers performance improvements with training time, even with limited data

HPC RESOURCE
MANAGER
APP FACTORY

Cloud-native API support and rapid prototyping allows for iterative design and deploy cycles for speed of integration into existing workflows

 

Tableau-based application deployment, for scalable, secure enterprise access

Platform
Application

Metonymize Applications

GAUGE

Helps marketers experiment by replacing real-world trial and error with simulation. 

  • GAUGE drives the marginal cost of market testing to zero from tens of thousands of dollars
     

  • With turn-around times of minutes versus days or weeks
     

  • Experimentation on GAUGE is granular, auditable and iterative in real-time

PING

Is an explainable language understanding solution for customer success managers

  • PING efficiently decodes customer behavior. Delivering insight with human-in-loop iteration
     

  • It delivers customizable, actionable data records from unstructured customer data.
     

  • Pretrained language models avoid “Cold Start” limits; 5x better accuracy with domain expertise

HOTLINE

Turn customer feedback into actionable data for operations

  • For customer insight managers keen to drive action and NPS, Measurable and actionable bite-sized customer feedback​​

 

  • With API designed for easy consumption in operations workflow
     

  • Delivering Scalable realtime VoC access to individual department managers

How to get started with Metonymize?

DEMO
PROTOTYPE

Explore capabilities for your use case

Prove output with your data

PILOT

Interface with workflow (APIs, Web Apps)

ROLLOUT

Enter production environment

Marie Bernard
Law, Domain Advisor

Marie is CEO of NextLaw Labs, Denton’s Venture Arm. In 2017, Dentons was ranked the #1 International Law Firm

Team

Team

Over 20 years of hands-on research and early-stage product development
Co-founder of EY's M&A Strategy practice, and lead Tech Partner

Kripa Rajshekhar
Founder, CEO

Led the text group on Watson Jeopardy!
Runs the Text Lab at University of North Carolina Charlotte
Visiting Professor at Duke University

Wlodek Zadrozny
NLP, Science

Two decades of experience, inventing applied machine learning solutions
Over 65 publications

Harsha Veeramachaneni
Applied Machine Learning

Senior Software Engineer, with a decade of experience in Cloud infrastructure, Clojure. Former CTO of Lumanu

Jake McCrary
Engineering

Advisory Board

Founder of AI Labs at both the University of Chicago and Northwestern University. Cofounder of Narrative Science

Kris Hammond
Senior Advisor

She also serves on the Board of Directors at Corning Inc., Meritage Homes, and American Eagle. Previously President of P&G’s $20B Global Beauty Sector, and enterprise eCommerce

Deb Henretta
Senior Advisor

Jim leads business development in Consumer Products & Retail, with over 30 years of experience in the industry, most of them at Procter & Gamble

Jim Prevost
Senior Advisor

Former Chief Technology Officer at The Hershey Company and Chief Information Officer at Chiquita Brands

Waheed Zaman
Senior Advisor
Credentials

Credentials

Backed by leading research and academic institutions, the team has a collective 100+ years of AI experience and over 100 business transformation projects at some of the most respected firms on the planet (including, IBM, P&G, EY, and Mondelez)

  • “Mined Semantic Analysis (MSA)” patent, for understanding granted by the USPTO
     

  • Collaborated with Harvard Business School on a study of CPG innovation and AI 

  • White LinkedIn Icon
  • White Twitter Icon
  • White Google+ Icon
  • Mined Semantic Analysis (MSA) presented at AAI ’15 (World’s leading AI conference)
     

  • Initial funding from UNC Venture incubator 
     

  • First Patent Filed

  • White LinkedIn Icon
  • White Twitter Icon
  • White Google+ Icon
bottom of page