Rainbird is an AI-powered automation platform that is made up of two crucial parts:
an easy-to-use modelling interface
a powerful reasoning engine
Your Subject Matter Experts (SMEs) use The Editor to encode their business logic and create knowledge maps.
SMEs can literally draw their thought processes including concepts, relationships, facts and rules to create a scalable and flexible repository of knowledge.
With its user-friendly interface, Rainbird is designed to be used by business-people, not tech teams.
The Editor is intuitive so your team can publish solutions fast, whether they’re integrating into your current systems or creating a new product, such as a consultative chatbot.
Users can choose to model their knowledge visually, switch to our in-line modelling language (RBLang) or toggle between the two. RBLang is easy to learn, and includes a full range of mathematical operators, and easy connectors to external data sources and third-party APIs.
Giving access to RBLang also makes it easy to import structured knowledge from other formats, enabling you to rapidly build your Rainbird project.
The Runtime is the brains behind the operation, the powerful decisioning engine that uses the models created in The Editor to automate decision-making. It is the engine that you will never see.
Rainbird looks at what it knows before making a decision or recommendation. If the Runtime doesn’t have all the data, or is unsure about something, it may ask the user a question (or series of questions). We refer to this process as the Match, Infer, or Ask (MIA) methodology.
It asks the fewest questions of the user that will result in the highest certainty output, while keeping the conversation as human-like as possible. It also manages concurrent queries made by users, whether via web assistants, chatbots, apps or other worker applications.
The Runtime is either hosted as part of our enterprise cloud solution, in your own cloud or on-premise behind your firewalls. Rainbird is easy to deploy and manage in most data centres.
- It can 'reason'
Rainbird’s flexible rule engine can cope with uncertainty and missing data by ‘connecting the dots’ to answer queries. It does this by enabling you to define individual nuggets of knowledge as rules throughout your Knowledge Map. These rules can express uncertainty; so you can state how convinced you are about the rule itself, and how salient individual conditions are. This combination enables you to efficiently create exceptionally nuanced models that can cope with the fact that life is rarely black and white.
- It links knowledge
The platform drives smarter decision-making by enabling semantic links between different Rainbird knowledge maps. This enables you to 'join-up' previously siloed knowledge and deliver a more holistic and strategic system, capable of automating complex decisions.
- It provides evidence
Rainbird provides an audit trail for every decision – which is a key differentiator to machine learning. In today’s compliance-driven world, you cannot automate important decision-making without being able to articulate why a judgement was made. Rainbird’s audit trail makes it an ideal choice for regulated industries. Rainbird can explain how every judgement has been reached, including its certainty and every factor that went into making that conclusion, as well as where it sourced the data it has assessed.
- It interacts like a human
When your users consult with Rainbird, its artificial intelligence engine ensures that subsequent questions are considered in context, just like human interactions.
- It is open architecture
Rainbird Knowledge Maps can be quickly connected to external data sources. This means you can reach out and grab the real-time data needed to feed into an automated decision-making process. It’s also simple to connect to any third-party API, enabling your model to seamlessly integrate with other third party systems. For example, you can build a credit scoring model that pulls client data from your CRM, data from external credit agencies and applies your own model of automated decision-making on top. What you build in Rainbird you can also publish as an API for easy integration into other software.
- It has 'One click' publishing
So you can quickly publish a chatbot that anyone can use to diagnose problems, obtain advice or solve problems. You can also embed a web agent in any website using a single line of code.