Using Deep Learning for On-demand Expert Service Cloud for Analytics

Print Friendly, PDF & Email

Rumblings in the industry indicate there is a new on-demand expert service cloud for analytics. BI teams can now tap thousands of experts around the globe to quickly handle ad-hoc queries and eliminate backlog. The new technology creates a pool of analytics experts to add BI team capacity for ad-hoc analytics on data warehouses.

What’s the big deal? 

Analytics databases and software tools offer users tech support, but most users actually need to engage with experts to help them use the best product features and have successful in-product experiences to get their job done. BI teams need to quickly build out and train machine learning systems with the semantic understanding needed to correctly respond to any user query, giving product and marketing teams the self-service analytics they crave so they can continue to grow the business.

Case in point, Got It, an on-demand expert platform and artificial intelligence (AI) innovator, launched Querychat for cloud data warehouses starting with Google BigQuery. The first on-demand Expert Service Cloud for analytics, Querychat creates a scalable and elastic pool of vetted analytics experts to add business intelligence (BI) team capacity with usage-based pricing, and automatically delivers deep learning training data into a natural language processing (NLP) AI system. BI teams can now tap thousands of Got It experts around the globe 24X7 to quickly handle requests for queries and dashboards and eliminate backlog, while labeling user questions and corresponding SQL queries to train a Google BERT-based, Querychat TrueNLP AI without compromising data privacy.

“Analytics tools and cloud data warehouses are extremely sophisticated products. We help users get ten times more value from these products by empowering users to collaborate with our Querychat analytics experts anytime – just as we do for Microsoft Excel and Google Sheets users with Excelchat. Our on-demand platform with an analytics Expert Service Cloud now also gives line-of-business users and BI teams access to vetted analytics experts within 30 seconds for a 20-minute, chat messaging workflow to help them create the right query or dashboard,” said Peter Relan, chairman and CEO, Got It. “BI teams can now scale easily to handle unlimited requests, via a low-cost, usage-based pricing model. Plus, we are solving the single biggest problem in machine learning: quickly getting the training data needed to build and train a company schema-specific AI model.”

Got It takes advantage of the advanced, full-sentence NLP in Google BERT, which is already pre-trained on a massive English corpus of 3.5 billion words. Leveraging BERT transfer learning for NLP and the data exhaust from user-expert chat conversations to train Got It’s deep learning AI model, Querychat breaks the constraints of keyword search-based, auto-complete Q&A available today.

“BI and analytics teams are always strapped for time and resources, making it hard to address all the queries users have of their data,” said Alec Winograd, founder of AppsCanvas. Wingrad, who built the Google BigQuery data warehouse for Quizlet, the leader in learning tools for tens of millions of learners, continued, “When you’re logging almost 10B events a month into BigQuery, you can save a lot of time and money with a system that protects data and handles natural language queries accurately, combined with an Expert service to add analyst capacity on-demand.”

Big Steps for Google BigQuery & the Columnar Cloud

Either from the Querychat TrueNLP AI or an on-demand Querychat expert, BI teams using columnar data warehouses in the cloud now have a way to guarantee answers to questions instantly, empowering marketing, operations, business and product managers to ask ad-hoc questions in plain English sentences. By only providing access to the data warehouse schema and not relinquishing control of or access to the data, companies are able to ensure data privacy while avoiding the long installation and configuration time issues of existing solutions. BI teams also get a workflow to verify Querychat results.

Currently, Querychat supports Google BigQuery cloud data warehouses carrying Google Analytics and Salesforce CRM data. Got It plans to extend Querychat support to Azure Cloud Data Warehouse and other environments in 2019.

Contributed by Daniel D. Gutierrez, Managing Editor and Resident Data Scientist for insideBIGDATA. In addition to being a tech journalist, Daniel also is a consultant in data scientist, author, educator and sits on a number of advisory boards for various start-up companies. 

Sign up for the free insideBIGDATA newsletter.

Speak Your Mind

*