Knowledge-Based Process Automation.

Self-learning AI understands, indexes and connects all your business data enabling new levels of automation

  • No software installation required
  • Innovative, industry-first pricing model
  • Supports text, audio, video and images
powered by metaphor.

What is Knowledge-Based Process Automation (KBPA)?

KBPA is a type of Business Process Automation that uses Knowledge Graphs (or Knowledge-Based Systems) and machine reasoning to accomplish complex tasks that humans can’t do. It requires understanding relationships between massive amounts of unstructured data in order to enable business decisions and insights.

What challenges are we solving for customers?


Challenge: Knowing which of your products are affected by regulation changes is critical, but daunting ― especially if you operate across multiple states and countries.

Solution: Automatic notifications when one of your products is affected by a regulation change. We’re able to scan updates and find links to your relevant products automatically. And our multi-language support makes sure you’ve got all your regions covered.


Challenge: Some customers have a large number of contracts they need to quickly extract information from to ingest into their SAP software. This data is unstructured and the information sought is complex, requiring an understanding of business clauses or terms ― not just pulling out names and numbers.

Solution: Automatic parsing and understanding of relevant business information, such as variable terms and conditions, across multiple languages. No large training sets required.


Challenge: One of the most common data challenges our customers face is a fragmented view of their business due to disconnected data silos. For example, having user emails, support chats, profiles, and transactions stored in different locations and not being able to work with this data holistically.

Solution: A centralized view of all this disparate data with links and relationships automatically generated ― whether the data is structured or, more commonly, not (raw text, audio/video recordings, PDF documents, etc). This enables deeper insight and powerful automation based on rules that can jump across data silos.


Challenge: The largest media companies produce tons of new content every day. Indexing, finding, recommending and creating this content is largely inefficient due to lack of automation.

Solution: We help some of the world’s largest publishers leverage the full potential of their content through automation. Automatic metadata generation from every piece of video, text and image content created allows for a fully-indexed library. This means content is easier to find for internal producers and external consumers. We’re also able to surface the most closely related content for any given piece by automatically uncovering relationships between items. Read more about media use cases in detail.

AI is complex ― pricing doesn’t have to be

We’ve caused quite a stir with our innovative pricing model for AI services.
It’s one of the many reasons people love working with us.

  • No setup fees
  • No model training or maintenance fees
  • Just one flat monthly rate from the start

Simple integration

1. Connect your data

2. Self-learning AI scans your content and finds relationships between all your information, regardless of where it lives or what format it’s in

3. Ready to go!
Visualize and automate