All communication is encrypted using AES 256 bit encryption with support for perfect forward secrecy (PFS). We have taken special care to ensure all communication is secured without compromising the high-speed delivery, low latency and reliability on weak connections.
Avaamo services are run on hardened Linux VM’s. Data storage and processing are performed on servers that cannot be accessed from the internet. We also utilize technologies like certificate pinning to control client access to all production machines. All production machines are audited periodically as well.
Avaamo supports multi-factor authentication using the one time password (OTP) technology to validate user’s device identity as well as corporate identity before providing access. Avaamo can also optionally support additional corporate authentication mechanism such as SAML 2.0 based SSO and OAuth2 for direct authentication against Active Directory or any LDAP compatible server or Identity Management provider.
Avaamo’s infrastructure is already certified for PCI DSS Level 1, HIPAA, ISO 9001, ISO 27001, SOC 1/ ISAE 3402, SOC 2 Type 2, and more. Details on all the infrastructure certifications are available upon request. Avaamo also utilizes several third-party partners to perform various certifications ranging from penetration testing, mobile vulnerability testing, MITM attack tests, Dynamic Application Security Testing (DAST), Static application security testing (SAST).
Want to know the overall accuracy of the bot over the last months, trending intents of the week, overall user sentiment on the bot, which channels have better conversion rate – you can get all this information and more from the Analytics section of the dashboard. Get a unified view of access and usage across all channels and usage trends across timelines. Drill down specific to an intent or domain or
knowledge pack to see how effective the bots are in effectively resolving and reducing the time to resolution of queries. You can also track user sentiment across various intents and track the overall bot usage sentiment. Using the agent console you can also get a real-time view of user queries and bot responses and also “take over” any of the conversations.
All user queries that the bot could not respond to or gave incorrect responses across all channels are centrally collated in the unhandled Message Queue. The entire conversation context, tone, closely matching intents and suggested intents are made available to the bot developer to assist them to quickly train the bot to handle these queries.
Various types of user feedback and survey forms are available as templates that can be enabled as part of any of the conversation flows. User feedback via a simple Like button can also be triggered as part of any response and the feedback is utilized in real-time to train the bot and present relevant options to users. You can quickly create NPS reports based on user feedback as well as export to external analytics tools as well.