Petuum is on the leading edge of making AI accessible to enterprises of all size and stripe. We aim to empower the users of our technology to solve the most difficult real-world problems. We’ve recently posted a recording from our Digital Networking webinar series showcasing a series of technical demonstrations of AI technology that can be applied across a wide range of uses:
Petuum offers solutions at every level of deployment process to quickly develop and implement solutions at scale on commodity hardware useful to enterprises building AI solutions from the ground up using in house talent to those who are looking to quickly implement prebuilt solutions. This video quickly showcases a variety of solutions built by world class AI scientists in areas such as Computer Vision (CV), Natural Language Processing (NLP) and Complex Decision-making using real world applications that can be generalized to multiple industries. Demos include:
Petuum Optimum for Industrial AI
An end-to-end example of a Complex Decision product built on top of our AI platform, Orchestra, to enable AI-driven smart manufacturing. The process starts with building neural networks models based on data from multiple sources. Those models can then be used to predict outcomes as far as 30 minutes in the future based on the current state; suggest actions based on those outcomes; and even, depending on the level of integration, take semi-autonomous actions on the factory floor to implement those suggestions.
NLP Accelerator for Medical Records.
Using a solution developed for the medical field for ingesting and summarizing medical records, using a knowledge graph built from a broad range of sources from .pdfs to handwritten notes with very little human intervention.
CV Accelerator Combined with NLP for X-ray Analysis
Showing how CV and NLP can be combined analyze chest X-rays to quickly product detailed written reports in multiple languages, scaling for very large files and data sets.
Image Processing for Retail
Demonstrates how to streamline data preparation with key point matching to build sophisticated image analysis using small datasets from multiple sources.
Neurobots for Image Enhancement
A multi-step process that starts with identifying often difficult to articulate criteria about the quality of images which are then used to automatically improve existing or new images.
NLP Chatbot using multiple languages
An example of a chat bot that can process both multilingual datasets as well as natural language queries (English <> Japanese), that can be very quickly developed relying on keywords
All the demos include suggestions for how the specific solutions can be applied to other industries with a focus on industrial applications in areas such as inventory management or defect identification. These examples are but an initial sample of AI capabilities that can be incorporated by enterprises of all sizes to enable higher growths and lower costs.