What is Deflekt.ai?
Deflekt.ai is an AI tool that automates customer support email replies. It enables users to deflect emails before they reach the support inbox, allowing the support team to focus on real issues. It has the ability to work with an existing AI chatbot or use its built-in AI bot to automatically answer emails.
How does Deflekt.ai automate customer support email replies?
Deflekt.ai automates customer support email replies by utilizing either a user's existing AI chatbot or its built-in AI bot. When an email is received, it is stored for processing. The text content from the email is extracted, the customer's first name is parsed from the headers, and Deflekt.ai calls the chatbot API to get an answer. The customer then receives a reply with a disclaimer at the top.
Can Deflekt.ai be integrated with my existing AI chatbot?
Yes, Deflekt.ai can be integrated with your existing AI chatbot. It is designed to work with various integrations, allowing users to use their own AI chatbot to answer emails automatically.
What alternative AI providers can I use with Deflekt.ai?
In addition to your own AI chatbot, you can use alternative AI providers such as Mendable, AskAI, and Chatbase with Deflekt.ai. The tool is designed to support multiple integrations.
What is the purpose of the inbox alias in Deflekt.ai?
In Deflekt.ai, the purpose of the inbox alias is for automatic processing. Users can set up an alias for each inbox and a human destination. The inbox email address can be used publicly or to forward emails to. When an email is received, it is stored by Deflekt.ai for further processing.
How does Deflekt.ai handle data extraction and answer fetching?
Deflekt.ai handles data extraction and answer fetching by fetching the text content from incoming emails and attempting to parse the customer's first name from the headers. It then calls the chatbot API to get an answer along with source links, which is then sent as a reply to the customer, featuring a disclaimer at the top.
What is a fallback scenario in the context of Deflekt.ai?
A fallback scenario in the context of Deflekt.ai refers to situations such as deactivation of your inbox, reaching your Deflekt.ai plan limit, or if your chatbot API is down or you've run out of credits. In such cases, since the original message is stored, it can be forwarded to your human support inbox. This ensures no messages are missed.
What happens in Deflekt.ai if a customer replies to an incorrect AI chatbot answer?
In Deflekt.ai, if a customer replies to an incorrect AI chatbot answer, their response is relayed through Deflekt.ai for monitoring and ends up in your human support inbox right away.
What information can I monitor on the Deflekt.ai dashboard?
On the Deflekt.ai dashboard, users can monitor the number of emails ingested, those deflected (answered), skipped or undeflected (replies to answers). The dashboard showcases exactly how many emails were prevented from ending up in the ticketing system or support inbox.
Can Deflekt.ai reply in multiple languages?
Yes, Deflekt.ai can reply in multiple languages. Users have control over the reply email template strings and most AI providers that Deflekt.ai works with support multiple languages.
What plans does Deflekt.ai offer and what are their key features?
Deflekt.ai offers a range of plans. The free plan includes 1 inbox and 50 email deflections per month. The Starter plan at $9 includes 250 email deflections and 1 inbox. The Grow plan at $29 includes 1000 email deflections and 3 inboxes. The Pro plan at $49 offers unlimited email deflections and unlimited inboxes. All plans include all integrations.
Are there any refunds for canceled paid plans in Deflekt.ai?
Deflekt.ai does not offer refunds for canceled paid plans due to the nature of their product.
Is language detection a feature provided in Deflekt.ai?
Language detection is a feature that is currently in progress for Deflekt.ai. It is an upcoming feature that will aid in replying in the customer's language.
How does Deflekt.ai determine an 'email deflection'?
In Deflekt.ai, an 'email deflection' occurs when an incoming email is successfully replied to using AI integration. When the reply is sent, the email is considered as 'deflected'. If, for any reason, this fails, the email is marked as 'skipped' and it does not count against the monthly deflection limit.
What is an inbox within the Deflekt.ai system?
Within Deflekt.ai, an inbox is essentially an alias with configuration settings. In most cases, one inbox suffices. However, for users running multiple products, each with its own knowledge base and domain name, multiple inboxes may be beneficial as users can set from name, template strings, and human destination inboxes per inbox.
Can Deflekt.ai be tested for free before investing in a paid plan?
Yes, Deflekt.ai offers a free plan so users can set up an inbox, an AI integration and test out the platform. This free plan includes 1 inbox and 50 message deflections per month.
How do paid plans work in Deflekt.ai and how can I upgrade?
In Deflekt.ai, users start with a free plan and can upgrade to deflect more messages. The upgrade can easily be subscribed to within the app, and payments are handled by Stripe. Upon subscribing, users can cancel at any time, and their subscription will end after the subscribed period.
Can Deflekt.ai work without utilizing an AI chat tool?
Yes, Deflekt.ai can work without an AI chat tool. It has a built-in bot solution that works well for replying to emails, even though it does not offer a chat widget. This built-in bot is included in all plans, including the free plan.
Where can I find more detailed documentation for Deflekt.ai?
You can find more detailed documentation for Deflekt.ai on their website at 'https://docs.deflekt.ai'. This documentation includes a large number of guides, answers, and how-to's.
What future features are planned for Deflekt.ai?
While specific future features for Deflekt.ai are not mentioned, there is mention of features being developed based on user feedback. Also, language detection for incoming emails and having an inbox configuration per language is mentioned as an upcoming feature in the roadmap.