# How to Sign Up on Dowork.ai Creating an account on Dowork.ai is easy and efficient. Follow these steps to get started: **Step 1: Visit Dowork.ai's Landing Page** * Head over to Dowork.ai’s website. * On the landing page, you will find the "Try Now" button in two places: one at the center and another at the top right corner. * Click on either "Try Now" button to begin. **Step 2: Redirected to the Sign In Page** * After selecting "Try Now", you will be redirected to the sign-in page. * Here, you will see two main options for creating your account. **Step 3: Option 1 - Sign Up Using Email** * On the sign-in page, look for and click the "Sign Up" button. * This will navigate you to the sign-up page. * Fill in your email address and set a password. * Follow through any additional steps to complete the sign-up process. **Step 4: Option 2 - Continue with Google** * Alternatively, you can sign up using your Google account on the sign-in page. * Simply click on the "Continue with Google" button. * Choose an existing Google account or add a new one. * Follow Google’s prompts to complete the sign-up. **Need Help? Ask Our AI Agent!** * If you have any questions during the sign-up process, feel free to ask our AI Agent. * You can find the AI Agent in the widget on our website. It's there to help you with any queries you might have. Congratulations on signing up for Dowork.ai! If you face any issues, our support team is always here to assist you. Welcome aboard, and enjoy using Dowork.ai to boost your productivity! # How to Connect Your Website to Dowork.ai's AI Agent Connecting your website to Dowork.ai's AI Agent allows it to learn from your site's content, enabling it to provide accurate responses. Here's how you can set this up: **Step 1: Navigating to the Get Started Page** * After successfully signing up, you'll be directed to the "Get Started" page. * Click on the "Add Website" button. **Step 2: Adding Your Website** * A popup will appear where you should enter your website URL. * It’s recommended to use your website's root URL, like `https://dowork.ai/`. * After entering the URL, click the "Add" button. **Step 3: Monitoring the Crawling Process** * You’ll be taken to the setup page where you can see the status of the crawling process. * This page also shows how many pages have been crawled. * Once crawling is complete, the status will change to "Completed". * You will receive an email notification indicating that the crawling is finished and you can start testing. **Step 4: Testing the AI Agent** * Even before the crawling is fully complete, you can begin testing. However, note that the answers might be incomplete until crawling is finished. * Click the "Test" button to open a widget where you can ask questions and receive answers from the AI Agent. * This testing widget is similar to what users will see on your website. **Step 5: Integrating the AI Agent Widget into Your Website** * To integrate the widget into your website, click on the "Scripts" section. This will copy the required script to your clipboard. * Add this script to the `` section of every page of your website where you want the widget to appear. * If you're using a CMS, look for an option to add custom scripts. Otherwise, ask your web developer to do this for you. * Once the script is added, refresh your website to see the Dowork.ai widget. **Step 6: Updating Website Content** * If you update your website, click the "Recrawl" button. This ensures that the AI Agent learns from the latest content. With these steps, Dowork.ai's AI Agent will be successfully integrated into your website, equipped to assist users with information derived from your site's content. If you have any questions or encounter difficulties during this process, remember that our AI Agent is always available to help. You can ask any queries through the AI Agent widget on our website, ensuring you get timely and effective assistance. # How to Upload PDF Files for Extra Information on Dowork.ai For businesses that have additional information not available on their website, Dowork.ai allows uploading PDF files to further train the AI agent. This feature is especially useful for enhancing the AI's knowledge base with more specific or detailed information. Here’s how you can upload PDF files: **Step 1: Navigate to the 'Uploads' Page** * After logging into your Dowork.ai account, locate the 'Sources' section in the main menu. * Click on the 'Sources' section, and from the dropdown menu, select 'Uploads'. This action will navigate you to the 'Uploads' page. **Step 2: Accessing the Upload Feature** * On the 'Uploads' page, you will see a list of previously uploaded files, if any. * Locate the 'Upload File' button on this page. **Step 3: Uploading PDF Files** * Click the 'Upload File' button to select PDF files from your computer. * A file explorer window will pop up, allowing you to choose the file(s) you wish to upload. * You can upload multiple files by repeating this process for each desired file. **Step 4: Managing Uploaded Files** * After uploading, the PDF files will be listed on the 'Uploads' page. * Here, you can review, manage, or remove the files as necessary. Uploading PDF files on Dowork.ai enhances the AI agent's ability to provide more comprehensive and accurate responses, tailored to your specific content. If you encounter any issues or have questions during the upload process, our AI Agent is ready to assist. Simply reach out through the AI Agent widget on our website for help. # Configuring the AI Agent on Dowork.ai Dowork.ai offers a range of skills that can be configured for the AI Agent to cater to specific needs of your business. Here’s how you can select and configure these skills: **Step 1: Accessing the AI Agent Page** * Log in to your Dowork.ai account. * From the left menu, navigate to the "AI Agent" page. **Step 2: Selecting the Desired Skill** * On the "AI Agent" page, you’ll find options to select from three skills: General Assistance, Bug Troubleshooting, and Lead Qualification. **Skill Options:** **General Assistance** * Selecting this skill enables the AI Agent to generate answers based on the connected content. * Ideal for general queries and providing information. **Bug Troubleshooting** * If you choose this skill, the AI Agent will not only answer questions but also troubleshoot bugs reported by users. * It will try to provide workarounds or automatically open a ticket for unresolved issues. * To enable automated ticket creation, a ticketing system (like HubSpot) must be connected. * When selecting 'Bug Troubleshooting', a panel will appear on the right-hand side showing the ticketing system name and a 'Connect' button. * Click 'Connect', and you will be redirected to the ticketing system's page for authorization. * Once connected, Dowork.ai can create tickets automatically, streamlining the support process and reducing communication overhead. **Lead Qualification** * Choosing this skill enables the AI Agent to qualify leads automatically while answering user queries. * It identifies potential leads and automatically sends an email with the collected information. * This skill is useful for simultaneously providing information and capturing leads. Configuring the right skill for your AI Agent on Dowork.ai can greatly enhance your customer interaction, streamline support processes, and optimize lead generation. If you have any questions about configuring the AI Agent or any of its skills, don’t hesitate to ask our AI Agent for assistance. Just use the widget on our website for quick and efficient support. # Understanding User Interactions: The Conversations Section on Dowork.ai Dowork.ai provides a comprehensive Conversations section where you can monitor all interactions between users and the AI Agent. This feature is crucial for understanding how users are engaging with the AI Agent and for gaining insights into their needs and preferences. Here's how to navigate and utilize this section: **Step 1: Accessing the Conversations Page** * Log into your Dowork.ai account. * From the left menu, select the "Conversations" option. This will take you to the Conversations page. **Step 2: Exploring User Conversations** * On the Conversations page, you will see a list of users who have interacted with the AI Agent on the left-hand side. * Each user may have multiple sessions. Users are identified by their email and the country flag associated with their initial conversation. **Step 3: Viewing Specific Conversations** * By clicking on a particular session, you can view the full conversation between the user and the AI Agent. * This view also allows you to see which answers were liked or disliked by the users. **Step 4: Analyzing Conversation Summary** * At the top of the Conversations page, you can expand the "Summary" section. * Here, you’ll find key metrics like the total number of messages handled by the AI Agent and the number of liked/disliked messages. * Monitoring these messages is essential for understanding how users are communicating with the AI Agent. Regularly reviewing the Conversations section on Dowork.ai is vital for gaining a deeper understanding of user interactions. In the next article, we will delve into the "Common Queries" section for even more in-depth insights into user interactions. Remember, if you need assistance navigating the Conversations section or interpreting the data, our AI Agent is always ready to help. Just reach out through the widget on our website. # Analyzing Common Queries or Requests on Dowork.ai The "Common Queries" feature on Dowork.ai provides businesses with valuable insights into the most frequent questions or topics users ask the AI Agent. This information is crucial for businesses to understand user concerns and make strategic decisions. Here's how to access and use the Common Queries section: **Step 1: Accessing the Common Queries Section** * Log in to your Dowork.ai account. * From the left menu, select "Common Queries" to enter the section. **Step 2: Reviewing Automatically Generated Queries** * In the Common Queries section, you'll see a list of frequently asked questions or topics, automatically categorized based on user interactions. * Each conversation is linked to one of these common queries or topics. **Step 3: Gleaning Business Insights** * By reviewing these queries, businesses can gain a deeper understanding of what users are most curious or concerned about. * This insight can guide various aspects of business strategy, such as improving user experience, refining UI/UX, or addressing product quality issues. * For example, a high frequency of UX/UI related questions might indicate a need for better user interface design, while queries about bugs could point towards investing more in product quality and testing. **Upcoming Enhancements** * In future updates, Dowork.ai plans to introduce filtering options for these topics, allowing for more targeted analysis of user interactions. Regularly reviewing the Common Queries section is vital for businesses to stay connected with their user base and to understand prevalent concerns. This understanding can drive meaningful changes in your business strategy and operations. If you have any questions or need assistance with navigating this feature, our AI Agent is always ready to help. # Revolutionizing Data Extraction: Introducing Stun by Dowork.ai ### **Introduction** In today's information-driven age, the capability to effectively process and interpret vast quantities of unstructured text is a game-changer. Traditional models, such as Retrieval-Augmented Generation (RAG), have been instrumental in leveraging large language models (LLMs) for a variety of applications. However, they fall short in precision and specificity, especially in complex customer service scenarios. This is where Dowork.ai's innovative technology, Stun - Extracting **St**ructured Data from **Un**structured Text, steps in to bridge the gap. ### **The Challenge** Imagine a situation in a hospitality setting where a guest asks, "What are all the dining options in this hotel?" Traditional RAG mechanisms, which primarily vectorize text chunks and retrieve the top few results from a vector database, often return incomplete answers. This approach, while effective for certain queries, struggles with more nuanced requests. It's particularly problematic in cases that demand detailed aggregation or specific data points, such as identifying the least expensive item on a menu or the restaurant with the most diverse cuisine options. ### **Introducing Stun** Dowork.ai's Stun technology heralds a new era in data extraction. It fundamentally changes the approach to processing unstructured text by: * **Entity Extraction**: Employing advanced LLMs, Stun meticulously extracts entities from raw text and formats them into a structured JSON document. This process involves a nuanced, multi-layered algorithm that not only identifies but also categorizes data with high precision. * **JSON DB Queries**: When a query is received, Stun doesn't just search for answers; it constructs a JSON-based database query. This enables targeted data retrieval, ensuring that the responses are not only accurate but also relevant to the specific query. * **Response Generation**: Leveraging the structured data, Stun then crafts responses that are not only precise but also contextual and comprehensive, a stark contrast to the often fragmented answers provided by traditional RAG systems. ### **Benefits of Stun** Stun stands out in several key areas: * **Precision in Data Retrieval**: By converting unstructured text into structured, queryable data, Stun ensures that the responses are complete and encompass all relevant information. * **Enhanced Explainability**: Unlike black-box models, Stun offers transparency in its responses, making it easier for users to understand how and why a particular answer was provided. * **Adaptability and Scalability**: Stun's algorithm can adapt to various domains and scales, making it suitable for diverse applications, from customer service to market research and beyond. ### **Real-World Applications** * **Customer Service**: In sectors like hospitality, retail, and banking, Stun can provide detailed, accurate information to customer inquiries, enhancing user experience. * **Data Analysis**: For researchers and analysts, Stun can sift through extensive documents, extracting and structuring key data points efficiently. ### **Conclusion** Stun by Dowork.ai is not just an incremental improvement but a transformative approach to handling unstructured text. Its potential to reshape industries through precise, explainable, and scalable data extraction is immense. As we advance into an increasingly data-centric world, Stun's role in harnessing the true power of unstructured text becomes ever more critical. ### **Demo** {% embed url="" %} Hi everyone, in this video, I'm going to demonstrate our new technology called Stun, which is a new RAG mechanism that drastically improves the quality of generated answers when building AI agents. I'll explain how traditional RAG retrieval falls short in certain cases and introduce our new approach that uses JSON databases to query specific entities and generate accurate answers. I'll also show you how to use Stun to crawl a website, extract entities, and build a JSON database. Join me to learn more about this exciting technology! {% endembed %}