Have something to say?

Tell us how we can make our product better.

Pinned

Completed

FAQ: Questions and Answers

Q: What is FineGuide.ai and how is it different from other products? A: FineGuide.ai provides GPT-based highly customized chatbots trained on your data (website links, documents etc.). Our mission is to make customer interaction with your website a great experience. Our assistants are highly accurate thanks to our proprietary data processing algorithms on top of plain GPT output. We also provide a wide range of additional features like widgets, no-code assistant personality tuning, tracking and analysis and many more. This combined with a modern responsive UI that can be easily integrable into any website. Q: How do I know if your solution is suitable for my business? A: Traditionally businesses that faces a lot of customer interaction will benefit the most. This is because human customer support is expensive. Using AI support can reduce from 70% to 90% of your interactions with customers, and its also real time! But in more board terms any business could benefit from an interactive support based on AI. Q: How the assistant reduce the hallucination rate in its responses? A: We use special techniques developed in-house to structure the input data in a graph like structure. We also apply a form of summarization with a specialized indexing to locate the relevant information better based on the user input. As a result we are able to deliver fast and highly accurate responses. Q: Is it possible to customize the chatbot's personality? A: Yes, you can totally customize the chatbot's personality to match your brand's voice and values. This way, the chatbot can interact with your customers in a way that resonates with your brand. From the Assistants setting page you can fine tune a number of parameters that will affect its behaviour. In addition to that you can fine tune the system prompt for advanced use cases. Q: Can the chatbot handle complex queries effectively? A: The chatbot is designed to handle complex queries effectively by leveraging advanced algorithms like data structuring and context optimization. On top of this we also support actions which can perform specialized queries on behalf of user. Q: Do you support chatbot actions? A: Its is possible to use advanced actions. This feature is currently under development and available to early adopters. Q: What kind of customer feedback does the system collects? A: All the conversation logs are available to the chatbot admin, and is available on the website. We do offer analytics and advanced reporting as a pack of our premium packages. Q: How secure is the chatbot in handling sensitive information? A: We take privacy with high responsibility. We do not track users in any way. The conversation logs are only available to the chatbot administrator. We do not use data collected by the users in any way. Q: Can the chatbot integrate with other platforms or systems? A: The chatbot can be seamlessly integrated with various platforms and systems to enhance its functionality and provide a more streamlined experience. Whether you're looking to integrate it with your website, CRM system, or any other platform, the chatbot can adapt to suit your needs. Currently we have out of the box integration with Zendesk, Helpscout. For the widgets we can integrate with Calendly, Youtube, TradingView for financial data. Many more platforms will be added soon. Q: Does the chatbot support multiple languages? A: Definitely, the chatbot support most of the languages. You can also restrict it to some specific language or allow any language which is by default. If your training data is in another language the chatbot will translate it and respond in the user's language. Q: What kind of training data is used to improve the chatbot's responses? A: The training data used for the training must be provided by the chatbot admin and the configuration phase. You can provide website links or upload documents. We do not collect user conversations for training the chatbot. Q: How quickly can the chatbot learn from user interactions? A: This is a planned feature, however its usage is optional. It not might be suitable in all of the cases. We provide a way for users to provide a feedback by using thumbs up or thumbs down buttons. This information could be helpful to reinforce the assistant knowledge about the best answers. Q: Does the chatbot have a limit on the number of users it can support simultaneously? A: The number of simultaneous session is not limited. However the number of messages are limited by the package. Q: What kind of maintenance or updates does the chatbot require? A: The chatbot does not require any assistance or update. Since its a SAAS solution all the infrastructure is hosted on the provider side. The updates and new feature therefore are updated to everyone as soon as they are rolled out. Q: How does the chatbot handle user queries that it cannot answer? A: There are a few scenarios. It is possible to escalate to a human operator, or it is possible to create a ticket for the business owners to follow up. Other more advanced scenarios are possible thru action integration. Q: Is there a demo or trial available to test the chatbot's capabilities? A: Yes, there is a Free package for beginners, it is completely free of charge with a limited number of conversations. This hands-on experience will give you a better understanding of how the chatbot can benefit your business and engage with your customers effectively. If you want to check out advanced features you can get in touch with our sales team. If you have any other questions please add them in the comment section below and we will add them to our FAQ.

fineguide 10 months ago

💡 General

Completed

Adjust Asisstant answers with Q & A

Description: We want to enhance the learning material for each bot by adding a question and answer table. This feature will allow users to identify incorrect answers by reviewing the conversation history. Users can then click on a link to open a popup with two fields: the question and the bot's answer. Both fields will be editable, allowing users to adjust the question and the answer. Upon saving, this information will be stored in the bot context, enabling the bot to use this Q&A for future interactions. Functionality: Identify Incorrect Answers: Users review conversation history to find incorrect answers. Open Popup for Editing: Click on a link to open a popup with two fields: Question Field: Pre-filled with the question from the conversation. Answer Field: Pre-filled with the bot's answer. Edit Fields: Users can edit both the question and the answer. Save Information: Save the edited information to update the bot's context for future use. This feature aims to improve the bot's accuracy and learning capabilities by leveraging user feedback.

fineguide 3 months ago

💡 General

Completed

Multilanguage Support

We’re thrilled to announce that our chat window now supports multiple languages! You can select your preferred language for the chat interface, making it easier and more convenient for everyone to communicate in the language they’re most comfortable with. How to Get Started: Simply head over to the “Appearance” tab on your Assistant page and choose your desired language. The chat interface will instantly update, providing you with a more personalized and user-friendly experience. Supported Languages: Czech (cs) Danish (da) German (de) Greek (el) English (en) Spanish (es) Finnish (fi) French (fr) Hungarian (hu) Italian (it) Japanese (ja) Dutch (nl) Norwegian (no) Polish (pl) Portuguese (pt) Romanian (ro) Russian (ru) Slovak (sk) Swedish (sv) Turkish (tr) Ukrainian (uk) Vietnamese (vi) Chinese (zh) We hope this new feature enhances your experience with us. If you have any questions or need assistance, our support team is here to help! Thank you for choosing us, and happy chatting in your preferred language! 🎉 Best Regards, Fineguide Team.

Liviu Maftuleac 3 months ago

💡 Feature Request

Completed

Roadmap

Dear friends, we’re happy to share with you our product roadmap. Here are some exciting developments we have in the pipeline: 1. Instant Promoting Interface Without the Need to Create a Chatbot Target: October 2024 Reasoning: Capture user segments beyond B2B Description: Develop a chat interface similar to OpenAI's ChatGPT that allows users to start a conversation, drop documents, links, and set rules within the chat. Detailed Sub-features: Quickly start a new chat from scratch. Drop documents and links for crawling in the chat. AI remembers and applies user-provided instructions. Save assistant configurations as "workspaces" (assistant + rules + documents). Switch between workspaces. Export assistants to websites or other supported communication channels. Example Use Case: Lawyers managing client portfolios Load each client's documents, contracts, and reports into separate assistants. Manage multiple workspaces and switch between them as needed. Quickly create new workspaces without configuring the assistant. 2. Share Assistants Without Publishing to Websites Target: October 2024 Reasoning: Capture user segments beyond B2B Description: Allow users to share their assistants with others. Share via one-time invite link or email invite link, both with validity periods. Recipients enter their email to create an account and access the assistant. Sharing features are available to advanced tier subscriptions, but recipients can access for free. 3. Support Multimodality in Chat Target: End of 2024 Reasoning: Enhance user experience Description: Enable users to drop images and audio files in the chat, with AI capable of processing and responding to the media content. 4. Voice Chat Target: End of 2024 Reasoning: Enhance user experience Description: Allow users to communicate with the AI via voice, with AI processing and responding to voice input. 5. Multiple Voice Chat Engines Support Target: End of 2024 Reasoning: Enhance user experience Description: Enable users to choose between different TTS and STT engines, such as AssemblyAI for transcription and ElevenLabs for voice transformation. 6. Geo-Fencing Target: TBD Reasoning: Enhance user experience Description: Allow users to limit assistant access to specific geographical areas. 7. Access Control Target: TBD Reasoning: Enhance user experience Description: Enable password protection for shared assistants, allowing selective access. 8. White Label Solution (API and Chat) Target: End of 2024 Reasoning: B2B features Description: Allow companies to integrate assistants into custom solutions via API, with full control over assistant creation, learning materials, custom prompts, and more. Provide a chat endpoint and customizable chat windows for end users. 9. Integration with More CRMs Target: 2024-2025 Reasoning: B2B features Description: Add integrations with leading CRMs for lead collection, including Salesforce, HubSpot, Zoho CRM, Microsoft Dynamics 365, Pipedrive, Freshsales, SugarCRM, Nimble, Insightly, Oracle CRM, SAP CRM, Keap, Bitrix24, and ActiveCampaign. 10. Integrate with Zapier Target: End of 2024 Reasoning: B2B features Description: Allow users to build automation workflows using Zapier, adding custom steps in automation pipelines. 11. Support for No-Code Pipelines Target: Mid 2025 Reasoning: Enhanced user experience, B2B features Description: Enable users to build processes graphically without coding experience, featuring steps such as analyzing user intent, CRM lead collection, text processing and formatting, safety checks, and translations. 12. Internal Mini CRM Target: End of 2024 Reasoning: Enhance user retention Description: Provide an internal tool for lead collection, enabling users to manage leads without external CRM integration. Includes a minimalistic dashboard, status management, and CSV export functionality. 13. Multiple Communication Channels Target: Mid 2025 Reasoning: B2B features Description: Allow users to connect various communication channels directly to the platform, including Telegram, Viber, WhatsApp, Instagram, Facebook Messenger, Facebook Comments, SMS, and email. 14. Human-like Communication Capabilities Target: Mid 2025 Reasoning: Enhanced user experience Description: Allow AI to proactively engage with users, ask questions, and provide reminders, enhancing the interactivity of conversations. 15. Enhanced Chat with Reactions, Edit, Delete, and Reply Target: Mid 2025 Reasoning: Enhanced user experience Description: Add capabilities such as message editing, deletion, reactions, and replies to various communication channels and the native web chat interface. 16. Human Escalation Directly from the Platform Target: Mid 2025 Reasoning: B2B features Description: Enable operators to intervene in AI-human conversations, optionally taking over when necessary. Provide a UI for tracking and managing these conversations. 17. Custom Actions Target: End of 2024 Reasoning: Enhance user experience Description: Allow users to define and execute custom actions via API endpoints, with pre-defined actions like sending emails, human escalation, lead collection, and more. 18. NPS Score Collection Target: End of 2024 Reasoning: Enhance user experience Description: Enable clients to collect NPS scores at the end of conversations, with customizable messages. 19. Webhooks Target: End of 2024 Reasoning: B2B features Description: Provide the ability to send custom webhooks triggered by user messages or specific actions. 20. Enhanced Learning Context Manipulations Target: End of 2024 Reasoning: Enhance user experience Description: Allow users to view, edit, and manage the content they've added to the learning context. Enable users to preview website structures before scraping. 21. Specify Crawling Depth Target: End of 2024 Reasoning: Crawler features Description: Allow users to specify the depth of crawling, useful for targeted data collection. 22. Enhanced Analytics Target: End of 2024 Reasoning: Provide insights Description: Offer tools to analyze customer messages for insights such as topic trends, frequently asked questions, user sentiment, and more. Provide statistics like message volume, peak hours, geographical data, and demographic information. 23. Clone Assistants and Workspaces Target: End of 2024 Reasoning: Enhance user experience Description: Allow users to clone assistants and workspaces, including all configurations, rules, styles, and learning materials, for experimentation and modification.

fineguide 4 months ago

💡 General

Completed

New Chat Window Templates

Our latest update includes two more chat window templates. Now you can choose from three different templates depending on the assistant's purpose. Default Template This template looks modern and has plenty of room for messages and pretty touches. The font size is bigger, making it easier to read, especially for shorter replies. It's great for customer facing roles like landing pages, websites or e-commerce. Best suited for assistant roles such as Sales Assistants and Customer Service Assistants. Traditional Template This template blends basic decorations with clever space-saving design. It's perfect for assistants and platforms that need to fit into tight spaces, yet has to have a pleasant aspect. Minimalist Template The new Minimalist template is meant for space optimization. For this template we've elliminated most of decorative elements. It's best suited for platforms that have limited space. This template has a smaller font size and is space-optimized, making it more efficient for larger responses from the chatbot. The best scenario to use this template is for Knowledge-Base Assistants, where the answers might contain elaborated explanations and examples. Let us know in the comments which one you prefer. Feel free to suggest any improvements. Your feedback is highly appreciated. Thanks!

fineguide 8 months ago

2

💡 General

Completed

How to Integrate Fineguide with Helpscout

This document provides step-by-step instructions for setting up the integration with Helpscout. This integration enables users to transition from automated assistance to human support. Why this integration is usefull - while Fineguide handles most of the inquiries, some users might still need human interaction. This pairing becomes especially beneficial when most inquiries can be managed by AI, leaving only those that require human support for your customer service team to handle. Let's get started! Step I Login to your Fineguide account, go to Settings > Integrations > Helpscout click on the Copy the redirection URL icon. It's better to keep this tab open since we're going to need it in Step 3. Step 2 Open a new tab in your browser and log in to Helpscout. Once logged in, Click on Account > Your Profile From the left menu, click on My Apps > Create My App In the App Name write "Fineguide Chatbot", paste the callback URL from the Redirection URL from the previous step, and click Create. If all goes well, you should see a screen where you have App ID and App Secret. We'll need those two values for Fineguide. Step 3. Go back to Fineguide Tab, paste App ID and App Secret in their respective fields, and then click Submit. Follo the link to go to Helscout authorization screen. Click on the link to authorize Fineguide to send tickets to your Helpscout Inbox. After you click on the Authorize button, the screen should remain blank. Go back to Fineguide settings, and you should now see Helpscout activated. Next, click on the Mailbox you want to forward your tickets to. In my case, I have only one mailbox named "help". After you select and confirm the mailbox, you should see your chosen mailbox highlighted Step 4: Testing Now let's test if your tickets end up in Helpscout. Go to Assistants, click on any Assistant, and start a conversation. Once the conversation starts you should see a mailbox icon in the chat header. This should open a ticket creation pop-up. Complete the form and click on submit Once this is done, Fineguide should forward your Ticket to Helpscout Inbox. A copy of the ticket is also created in the Fineguide ticketing system. You should see the Helpscout logo if the ticket was successfully forwarded. That's it!

fineguide 9 months ago

💁 How To...