Panum Mindmelt™
Conversational AI (Virtual Agents) combining
natural language processing (NLP) with machine learning
Overview
Interaction between the government, individuals, and public and private organizations should work seamlessly and efficiently. In today’s world, this interaction can be enhanced through digital services that imitate human interactions. The Biden-Harris President’s Management Agenda (PMA) Priority 2, “delivering excellent, equitable, and secure Federal services and customer experience,” reflects today’s need for enhanced digital services. Panum’s Mindmelt Solution uses state-of-the-art Artificial Intelligence (AI)-based technologies, such as chatbots or conversational AI, to improve its customer experience. Panum Mindmelt™ processes large volumes of data through machine learning and natural language processing (NLP) to recognize speech and text inputs and translate their meanings across various languages. This imitates human interactions, offering the user personalized information, solutions, or recommendations based on their inputted search criteria.
Process
Conversational AI (Virtual Agents) combines NLP with machine learning. First, the application receives the information input from the human, which can be either written text or spoken phrases. If the input is spoken, Automatic Speech Recognition (ASR), also known as voice recognition, makes sense of the spoken words and translates them into a machine-readable format or text.
Second, the application deciphers what the text means. It uses Natural Language Understanding (NLU), which is part of NLP, to understand the intent behind the text. Next, the application forms the response based on its understanding of the text’s intent using Dialog Management. Dialog management orchestrates the responses and converts them into a human-understandable format using Natural Language Generation (NLG), which is the other part of NLP.
The application then either delivers the response in text, or uses speech synthesis, the artificial production of human speech, or text to speech to deliver the response over a voice modality
Last, but not least, the application accepts corrections and learns from the experience to deliver a better response in future interactions, which is the component responsible for learning and improving the application over time through the machine or reinforced learning.
Strategy
People are increasingly using conversational interfaces driven by choice, context, and convenience. Micro-conversations are 70-80% of all interactions in some industries that can now be fully digital across channels including the public sector. Panum provides a platform that allows the public sector to use multiple channels for interaction with vast amounts of information, requests/appeals, and services it provides for the public.
Capabilities
Pre-packaged
Domain-specific pre-packaged micro-conversation modules that can be easily stitched together quickly to create suitable conversation in certain business contexts in select domains.
Voice & Text Native
Support for a range of text and voice channels natively, including Alexa Google Assistant, WhatsApp, SMS Web Chat, FB Messenger, Slack, MS Teams, and more.
Free Form Human Like Conversations
Context Switching is a default part of all conversations, making conversations highly flexible. User initiated, bot guided, or mixed initiative conversations are a default part of all conversations.
Human Agent Handover
Incomplete conversations can be addressed by human agents. Agents can take control of conversations and engage with customers from a single inbox across various channels.
Domain Tailored Solutions
Domain trained and tailored solution for Front Desk Automation and Help Desk Automation in select domains
No or Low Coding
Limited coding required, only in the case of API integrations for new applications
Superior Technology
Many unique technology differentiators related to NLU – Intent Matching, Entity Detection and Extraction, Context Management, User Personalization etc
Deep Integration Partnerships
Deep integration partnerships allow for highly relevant and flexible dialog flows and readymade application integrations
Multi Language Support
Optimized for English and Spanish. It can be extended to other langauages.
Insightful Analytics
Extensive conversational analytics on leads, appointments booked, user interactions, etc.
