We stand with Ukraine

expert.ai Large Language Models and Knowledge Models

Smarter from the start with domain-specific context on the expert.ai Platform

Language is hard. Making sense of industry-specific jargon or company-specific product terminology and extracting additional insight at speed and scale is even harder. Until now.

With 300+ deployments of natural language solutions, expert.ai has developed deep domain expertise in insurance, technology, financial services, and media. The expert.ai Platform includes generative AI Enterprise Language Models and Knowledge Models which capitalize on that expertise with rules-based models that contain industry, role, or use case specific concepts and relationships that can be used out-of-the-box to quickly improve the accuracy of natural language results and surface new insights from unstructured data.

Make a project ‘smart from the start’ and provide every team member with intuitive and easy-to-use workflows and value that accelerate the development of custom and explainable natural language solutions.

Explore the expert.ai Enterprise Language Model for Insurance (ELMI) to add generative AI capabilities to a project.


Build Natural Language Projects Faster

Enterprise Language Models and Knowledge Models leverage out-of-the-box domain knowledge that provides a head start for building custom NLP solutions with the expert.ai Platform. Classify documents and extract relevant data for domain specific use cases with a high level of accuracy and explainability thanks to expert.ai’s hybrid approach.

Below is a sample list of currently available expert.ai Knowledge Models. New Knowledge Models are added everyday. Please contact us if you do not see a knowledge model you need to complete you natural language project.

Financial Information
Business Organizations

Identify and extract various types of financial organizations like central banks; banks; credit unions; ratings agencies; payment service providers; pension funds; insurance companies; brokerage institutions; publicly listed companies; etc.

Classify any kind of text that deals with the commodity market. This model focuses on the main sectors involved: agriculture, energy, livestock, metallurgy, and chemistry, and includes any related financial instrument, like derivatives and ETFs.

Classify any kind of text dealing with the topic of currencies and cryptocurrencies. Identify currencies, values and related trends described in text.

Identify and extract events and activities related to Finance such as Management and Ownership changes; acquisitions; downsizing; outsourcing; buy outs; etc.

Categorize text based on a taxonomy of economic terms and extract relevant data, such as the rising and falling of various rates and indicators. Use this topic to track events and trends in the world of economics.
Markets, Tickers, and Indices

Categorize content based on the geographical location of the stock market and extract entities related to stock exchanges; publicly listed companies; their corresponding stock symbols (i.e., tickers); and stock indices.

Categorize content about the securities markets, labeling texts according to the security type (i.e., Equity, Debt, Hybrid and Derivatives), and extract information about related entities, values and relevant trends.
Business Activities
Banking Customer Communication

Classify customer support emails and email threads to help sort and route 60+ of the most typical banking customer support communication topics.
Corporate Crime

Classify information about a company or its staff being involved in crimes or illegal acts to safeguard a business against corrupt customers or suppliers and monitor competitors.
Environment, Social, and Governance

Identify and categorize text for ESG-related concepts (Environment, Social and Governance), and categorize them as negative or positive based on the actions or statements attributed to a company, country, or institution.
Publishing & News Media
IPTC Media Topics

Media Topics is based on the IPTC global standards covering hundreds of news media subjects including sports competitions, political summits and elections, or breaking events such as military conflicts, terrorist attacks, natural disasters, etc. for classification.
Life Sciences
Drugs, Diseases, and Symptoms

Extract various types of biomedical entities from text, creating a link with a controlled knowledge concept. This model integrates the Unified Medical Language System (UMLS) to extract entities belonging to three of the most common classes in the UMLS taxonomy: drugs, diseases, and signs or symptoms.
Sentiment, Emotions, and Hate
Behavioral Traits

Identify 72 types of personality traits—like curiosity, honesty, negativity, etc. to provide a more human-like understanding of the content.
Emotional Traits

The Emotional Traits classifier can identify 39 feelings in text to better understand the emotions, opinions or attitudes conveyed in a text.
Hate Speech

Find instances of offensive and violent language in text and categorize them into different hate speech categories like personal insult; racism; sexism; ableism; religious hatred; classism; and body shaming.
Personally Identifiable Information (PII)

Identifies Personally Identifiable Information (PII) in text. PII is any data that could potentially identify a specific individual, including names, date or place of birth, phone number, physical address, and more.
PII with Pseudonymization

Identifies PII in text and returns the text with the identified values anonymized as pseudonyms.
PII with Redaction

Identifies PII in text and returns text with identified PII data redacted as “X” characters.