
Mendable by SideGuide is an AI-powered chat application tool that simplifies the deployment process for businesses looking to integrate smart search functions and AI chatbots. It is designed to automatically answer technical questions by training the AI on organization documentation, reducing the workload on teams. Mendable offers features like continuous model training, enterprise-grade security, multiple base models, and seamless integration, ensuring accurate, secure, and up-to-date chat applications. The platform supports various use cases such as documentation search, sales enablement, and product copilots.
Some key features of Mendable include:
Mendable is not open-source but offers some open-source components and integrations. The pricing for custom plans is based on different factors like use case, total usage, and custom work requirements. Mendable also provides a free plan with a limit on the number of monthly chats and offers free access to certain open-source projects upon eligibility confirmation.
Mendable was created by the company SideGuide, with details available about Mendable's features and functionalities. Mendable simplifies the deployment process of AI-powered chat applications, allowing businesses to integrate smart search functions and AI chatbots effortlessly. The platform supports multiple use cases like documentation search, sales enablement, and product copilots, providing features such as continuous model training, enterprise-grade security, multiple base models, and customization options.
To use Mendable, follow these steps:
Connecting Data: Start by connecting your data to Mendable through a simple online GUI or API. Easily manage different data sources by adding, modifying, or deleting them.
Teach Your Model: Customize your model by editing base model properties. Mendable supports models like GPT-3.5-Turbo and GPT-4. Train the model by correcting its answers and editing prompts.
Continuous Training: Improve AI accuracy over time by correcting wrong responses. Mendable offers training through answer correction and allows model creativity control.
Data Synchronization: Keep your data updated with Mendable's reingestion process, using CRON jobs and webhooks to ensure data sync and accuracy.
Customer Support Integration: Redirect customers to your support link when the bot can't answer questions. Mendable provides custom private and open-source models for various needs.
Customization: Customize your model to fit your company's needs through various features. Mendable supports multiple base LLM models, training through answer correction, custom prompt edits, and model creativity control.
Deployment: Deploy Mendable quickly with one line of code, internally or externally. The platform offers insights into user queries, interactions, and product usage, allowing for real-time monitoring and optimization.
Remember, Mendable offers a free plan and enterprise-grade security features. For more details, refer to the Mendable documentation.
I appreciate the wide range of data connectors that Mendable offers, making it easier for us to integrate with tools we already use, like Salesforce and Google Drive.
The user interface feels a bit outdated, and it sometimes takes longer to train the AI on specific documents than I anticipated.
Mendable significantly reduces the time our customer support team spends answering the same technical questions repeatedly, allowing us to focus on more complex issues.
I love the automated reingestion feature that keeps our data up to date without manual intervention. It saves us a lot of time.
While the platform is powerful, the initial setup can be overwhelming due to the number of features available.
Mendable helps us streamline our documentation search, which has improved our response time to customer queries, enhancing overall customer satisfaction.
The customizable model training feature is fantastic. It allows our team to correct answers and improve the AI's responses over time.
The documentation could be clearer, especially for new users who are not familiar with AI chatbots.
It helps us maintain a high level of support efficiency by ensuring that the AI can handle a significant number of common queries without human intervention.