There is a wealth of information tied up in text documents from entire files to individual tweets. Organizations have been extracting information from unstructured text for decades, but until recently they've been limited to keywords and named entities like known locations and names, which leaves a substantial amount of knowledge behind. Even today, extracting useful information from these text documents takes exorbitant amounts of time and manpower.

Jaxon automates and enhances the process of building custom classifiers so that what traditionally takes months is done in minutes. This allows businesses to access new domain-specific knowledge and automate time-consuming tasks.

Document Classification

Document classification is a massive issue for organizations due to the sheer volume of files generated daily by both employees and customers. Traditional approaches fall short in many cases, especially for domain-specific applications.

Voice of the Customer

It is crucial for businesses to understand what their customers are saying about the business, the products, or personal preferences in general. Customers are continuously providing this data via sources like reviews, forums, call/chat logs, social media, and satisfaction surveys. Jaxon allows businesses to use this information to:

  • Programmatically update product descriptions
  • Recommend resolutions for customer service
  • Triage logistics issues
  • Predict and prevent returns

Voice of the Patient

Valuable information about patients, their diagnoses, and their symptoms are recorded in doctors’ notes, social media posts, and medical forums. This information is typically ignored as the text is unstructured and hard to use. Jaxon can extract much from these sources, allowing businesses to learn crucial information for use cases such as:

  • Pharmacovigilance
  • Patient identification
  • Personalized medicine
  • Clinical decision support

Voice of the Worker

Workers often take freeform text notes detailing relevant observations and exactly which tasks they performed. Typically overlooked, these notes carry a lot of useful information. The amount of historical data usually available also makes these ideal for AI analysis and can be helpful for:

  • Morning reports
  • Preventative maintenance
  • Sales churn analysis
  • Investigations
  • Compliance

Process Automation

By learning from historical data and being able to ingest new data in real time, Jaxon is ideal for helping companies implement AI and ML to automate business workflows. Jaxon can train applications to extract information from unstructured text or classify documents allowing the data to be used as quickly as possible after it has been created.

Triaging Trouble Tickets

Jaxon understands the unstructured and semi-structured text present in trouble tickets and classifies them to ensure that the most pressing trouble tickets receive human attention first. Jaxon’s speed also allows new problems and patterns to be identified as quickly as possible for prompt response.

  • Accurately classify trouble tickets
  • Ensure correct automatic routing
  • Analyze context and historical patterns

Claims Adjudication

Jaxon rapidly and autonomously classifies insurance claims, eliminating the need to manually read this text. This then allows a downstream model to quickly flag claims that require human intervention and ensures that agents’ time is being put to the best possible use.

  • Identify missing information
  • Review against existing policies
  • Contextualize with historical data and records


Customers expect 24/7 customer service and most companies now use chatbots. The better and more natural-sounding the chatbot is, the more satisfied the customer will be. Given the nuances of natural language, training a bot to sound like a human is a feat, but with Jaxon, top-of-the line domain-specific chatbots are easy to train and deploy.

Intent Classification

Jaxon trains natural language understanding (NLU) models in support of chatbots and conversational AI systems. A typical NLU is a classifier that attempts to identify intents (what is a customer trying to accomplish?) and slots (easy to answer questions such as location) contained within user utterances. Training data, quickly provided by Jaxon, consists of sample utterances that are annotated to identify these intents and slots. Because Jaxon uses a company’s own data and does not require fully pre-labeled sample utterances, it is ideal for training domain-specific NLU models.


Contact Us


Jaxon is an AI platform that trains other AI to more accurately understand natural language from raw text.

contact us
(617) 506-9410
177 Huntington Avenue
Suite 1703
Boston, Massachusetts

© Copyright 2020. All rights reserved.

Close Menu