AI Virtual Assistants that Automates Enterprise Processes

Technology Overview

We build specialised AI virtual assistants for enterprises to automate conversations for internal and external customer support. As a market leader in Asia, we support multiple Southeast Asian languages to help companies scale up regional operations.

Our projects span across various fields, from sensitive industries such as BSFI, healthcare and government sectors, to eCommerce, logistics and education. Our virtual assistants help to alleviate functional departments of repetitive, monotonous tasks, like answering customer and employee enquiries and requests – they are more than your average FAQ machines as they go above and beyond to complete transactions and retrieve unique, personal information. 

Our customers have experienced up to 42% volume deflection of level 1, simple queries, allowing customer support agents to spend time on complex cases that add value to the top line. Most importantly, they have also witnessed up to 100-fold increase in response times, resolution rates and improvements in customer satisfaction levels. 

To meet enterprise client needs to serve different business units, our solution is able to authenticate different user groups and provide personalised answers based on their profiles. To empower smarter business decision making, we use analytics in the form intuitive diagrams along with dynamic sliding windows to visualize key insights, analyse qualitative and quantitative input of bot usage to derive customer behaviour and patterns.

Technology Features & Specifications

Proprietary Natural Language Processing (NLP) Engine

We use state of the art TensorFlow model as our primary deep learning framework. The neural network architecture for language understanding transformer applies a self-attention mechanism which directly models relationships between all words in a sentence, which allows our engine to understand even complex queries in a mix of languages. 

Entity extraction and unstructured data understanding

Able to perform semantic search with contextual understanding (not keywords) and to extract entities for more robust conversation transaction processing. This cuts down the number of intents to improve AI performance and reduce training effort. We are also able to perform language detection and to strip and censor sensitive information; all this can be administered and designed by non-technical users.

Bulk content management and training of chatbot

Allows admin users to easily upload, edit and manage intents and content to the bot to facilitate the continuous training of the bot.

Open & flexible architecture to facilitate transactions and integrations

Facilitates deep integration with existing CRM, knowledge base systems (e.g. SAP, Microsoft Sharepoint, Active Directory) for the end to end transactional capabilities, and existing systems to authenticate users.

Detailed breakdown of chatbot and live chat analytics

Allows admin users to monitor bot usage and activity and to study analytics for different time periods using dynamic sliding windows to make business decisions. The details and metrics measured are completely customizable and dependent on the chatbot’s use cases and key functions. 

Live agent takeover

When the virtual agent is unable to handle the question, it can be escalated to a live agent on our platform. This allows for a seamless transition and facilitates knowledge transfer and management support. 

High enterprise security measures and cloud/on-premise deployments

Hosting compliant with ISO27001 and world-class standards. Vulnerability and penetration testing, and real-time anti-malware monitoring. Multi-factor authentication and role-based access. Data encryption methods include: TLS/SSL, IPSec, AES-256, NIST, BitLocker encryption & TDE. Flexible deployment options for security compliance. 

Potential Applications

Financial Services, Insurances and Banking

  • Insurance calculator to help individuals calculate exact coverage and premiums they require based on their parameters
  • Handling the same queries from customers and financial consultants/agents with different sets of answers using conditional content
  • Automating the claims process: application of claims, checking of claims status, checking of policy wording 



  • Triaging of symptoms
  • Helping patients find the correct specialist given their symptoms
  • Article search and subscription for healthcare tips and news
  • Booking of appointments with GP, Specialist and Health Screenings
  • Emergency escalation to call for an ambulance when the chatbot detects urgent situations
  • Clinic/hospital search based on location


E-commerce and Logistics

  • Handling common customer service and customer care questions
  • Transactional: real-time tracking of parcel status
  • Transactional: Rescheduling/cancelling/modifying delivery details
  • Displaying a catalogue of products in a carousel view
  • Customer profiling tool to help customers browse according to their interests


Human Resources

  • Onboarding virtual assistant to help new recruits assimilate into the workplace
  • HR Handbook assistant to answer HR policies related questions
  • PDF search tool to crawl through policy documents to return the best match answer even for untrained questions
  • Automated leave application and checking of claims status/leave balances

Market Trends and Opportunities

The global chatbot market size was 840 million US$ in 2017; of which, the Asia Pacific market is projected to witness substantial growth over the next decade owing to various technological developments across major economies. The regional market is expected to grow at a CAGR of 24.7 % from 2017 to 2025. This presents an opportunity for bot platform players like ourselves and similar, complementary customer experience solutions to enter the Southeast Asian market to service this growing demand. 

Customer Benefits

Businesses looking to increase efficiency and ROI in 2 ways:

  1. By automating repetitive, low dexterity tasks such as customer support and internal/external FAQs, and
  2. Automating common workflows for customers and employees. Ideally, these tasks should be of a sizable volume to require automation.

This allows businesses to place a virtual assistant (chatbot) as the first line of defence to handle L1 routine questions and transactions, freeing up time for the likes of customer service teams or human resource departments to focus on higher value-adding activities and enquiries that actually require human intervention and discretion.

Our biggest differentiator is in our ability to handle multiple Southeast Asian languages as well as a mix of languages (e.g. English and Bahasa Melayu in the same sentence). This allows enterprises that operate regionally to deploy one multi-lingual bot to scale across the countries.

Make an Enquiry