Testimonials

Testimonials

Testimonials V1

Striving to provide the best, out of the box, artificial intelligence solutions for small business.

Meet the Team
Testimonials V2

How it works

Ingest
Data Cleanse
Auto Learn
Deliver
  • Email files are exported from mailbox as EML, PST, OLM, EDB, EMLX or other file types.
  • Email files and their attachments are packaged and securely ingested through an encrypted pipeline to AWS cloud.
  • Files and attachments are tagged and placed into their S3 folders, respectively. They await data cleanse.
  • Emails are fetched from raw storage to begin cleansing process.
  • All personally identifiable information (PII) is removed.
  • Email signature data such as phone numbers, company logos, and redundant email addresses are removed.
  • Chain duplicates are removed so there aren't redundant answers.
  • Cleaned email data is sent to ebase.ai learning model.
  • Emails and attachments are sent through ebase.ai's Auto Learn Model.
  • The Auto Learn Model encodes the information for information retrieval.
  • The model puts the encoders into an index neatly.
  • As questions are submitted, they hit the Model Endpoint, which encodes the question for match in the index.
  • This allows information retrieval based on the users natural semantics, instead of keyword search.
  • User submits question through ebase.ai web application.
  • Question is sent to REST API via secure connection.
  • Question hits Model Endpoint for encoding.
  • Top 10 results are displayed in "most confident" order for user to view.
  • User is then allowed to thumbs up/thumbs down results to continuously train the model for better and better results over time.
Read the FAQs