AccelBot can meet your current and future business requirements with the following offerings:

• qBot – Answer questions from a corpus of text (FAQs)
• tBot – Performs transactions by integrating with back-end systems
• iBot – Manage business processes/logic involving multiple interdependent steps

ACCELBOT - UNIQUE FEATURES

• Multi-Agent Architecture - With multiple bots in the same application, it has a unique ability to manage human-like conversation without losing context E.g. During a sales signup dialog, general questions along with transactions do not derail the flow. Additionally, each bot can be scaled, based on the load.

• No Programming - Model complex business logic and workflow without programming which makes it easy to maintain.

• iBot - It can manage business processes/logic involving multiple inter-dependent steps. Task models can be modeled, built automatically from other data, or learned using machine learning. also, combinatorial complexity of tasks can be managed, with many choices and their inter-dependencies.

OTHER FEATURES

• Self-learning capability (supervised machine learning).

• Post user training, customers themselves shall be able to configure all the bots through modeling using yaml files

• Build once and deploy everywhere (Websites, Mobile, Inside Chat Apps, etc.)

• Analytics tailor-made, based on business requirements

• SaaS - On-premises deployment

IN-HOUSE BUILT TECHNOLOGY STACK(AI)

• AI Logic Engine - Model complex business logic and workflow without programming. User can make changes and revisions at any point in the process and info still relevant after a change will be re-used, so that the user does not have to re-start the process.

• NLP Engine - It allows a wide variety and flexibility in how people express their intentions, through spelling correction, stemming, use of synonyms/antonyms, etc.
  • Understand user intent using trained NLP. E.g. agree/disagree, help, revise, quit/restart, buy.
  • Extract meaningful entities from user utterances: E.g. task-specific entities.
  • Uses task models to train on entity extraction, Part-of-Speech tagging, and intents specific to the model.