Agents for everyone , for every task, made simple .

Alchemist is designed for anyone interested in AI agents and looking for an easy-to-use interface. It serves as a one-stop shop for language-based generative AI applications, focusing on productivity and targeting users seeking a comprehensive solution. Alchemist is particularly useful for individuals who want full control over their agent teams while keeping the interface simple and intuitive. With privacy as a primary concern, Alchemist is ideal for those who want to keep their data secure and private.

Elemental and Alchemist together form a complete solution, with Elemental serving as the engine for running agentic workflows and Alchemist providing an easy-to-use interface for designing, running, and monitoring them. From a technical perspective, Elemental is a Python library that may require programming knowledge to use, while Alchemist is an application designed to be accessible to everyone. Additionally, Alchemist offers extra features like document storage and an internal knowledge base, enabling out-of-the-box Agentic Retrieval-Augmented Generation (RAG).

Data privacy is a major concern for everyone, and Alchemist is designed to address this by running entirely on your system. When you choose local models, no information is shared externally. Your documents remain on your system, and agents access them locally. There is no hidden data sharing with Alchemist. AttoAgents does not collect any information about your agent teams, tasks, messages, documents, or any other parsed data. By using a local language model and a local embedding model (e.g., with the Ollama engine), you can be confident that your data stays private and secure.

Yes, Alchemist includes a built-in knowledge base that provides Retrieval-Augmented Generation (RAG) capabilities. You can create distinct collections of documents and upload them to Alchemist, assigning each to a specific collection. During the design phase, you can select which collections your agent will have access to. This allows you to extend your agent team's knowledge with your own documents and utilize them in your tasks. The documents are parsed, embeddings are created, and the data is stored in vector storage for semantic search. The specific RAG method implemented is the Agentic RAG.

Yes, Alchemist offers two subscription plans: 'Personal' and 'Business,' which you can choose based on your needs. If you are earning a salary for your work, you should select the 'Business' subscription. If you are using Alchemist for personal projects, the 'Personal' subscription is the right choice. Both plans are billed monthly and can be canceled at any time.

The answer depends on the language and embedding models you choose for your agent and knowledge base. If you choose local models, no data is shared externally. However, if you opt for cloud models, data is shared with the model provider, and you should refer to their privacy policy. For cloud models, the shared data includes the text of the messages you send to the agent, the documents you upload to the knowledge base, and the embeddings created from those documents. This data is shared with the model provider to generate embeddings and responses to your messages.

Every subscription to Alchemist begins with a 14-day free trial period. During the trial, you have full access to all of Alchemist's functionality. You can create agent teams, assign tasks to them, monitor their progress, and utilize the knowledge base. Additionally, you can upload documents to the knowledge base and organize them into collections. You can also use all available tools in Alchemist and run agent teams with them. Once the trial period ends, your plan will remain active, and you will be charged according to the subscription you selected. You can cancel your subscription at any time.

Yes, Alchemist, through Elemental language model module, can use Llama.cpp as a model provider. You can bring any model file compatible with Llama.cpp and use it in Alchemist. You can also work with local models that ensure your data is not shared externally. Additionally, you can use customized cloud models if the provider supports customizations. The choice is yours.

Since Alchemist uses Elemental as its agent engine, you can bring your own tools to Alchemist by creating a tool compatible with Elemental. For instance, you can create a Python script that uses Elemental's API to interact with the agent team. By developing a tool compatible with Elemental and placing it in the tools directory (accessible in the File section of Alchemist), you can incorporate it into your agent team. However, the tool will not yet be integrated into the Alchemist settings section and will need to remain self-contained.

You can cancel your subscription to Alchemist at any time. If you cancel during the trial period, you will not be charged. If you cancel after the trial period ends, you will not be charged for the next billing cycle. To cancel your subscription, go to the settings section in Alchemist and click the 'Cancel subscription' button. Once you cancel, your plan will remain active until the end of the current billing period, and no charges will be applied for the next cycle. Alternatively, you can cancel your subscription on this page by clicking the 'Cancel subscription' button after signing in.

Yes, you can run agent teams fully offline by selecting local models for your agent team and knowledge base. You can create multiple teams, allowing a local team to handle private tasks. The agent team will operate entirely on your system and will not send any data to the model provider. However, the Alchemist application will require an internet connection to verify the subscription status every 30 days.