Meet Friday’s Workshop

A workspace around the work

The Organizer groups projects, folders, and conversations. The Dashboard brings context and controls alongside the conversation instead of making every task start from an empty page.

Specialization with boundaries

Agent management, Skills & Tools, and model-management interfaces separate roles, procedures, authorized actions, and model selection. The options available to a user depend on their role and deployment.

Knowledge in context

Document workflows and memory scopes support different kinds of context. Verify what was extracted, retrieved, and supplied to the task before relying on an answer.

Creative workflows

Media Studio provides a dedicated interface for configured generation workflows. Available modes and performance depend on enabled services, models, providers, and hardware.

Start with guidance, then build a workflow

Use Friday’s Workshop for a small task before building a larger project. Information icons explain workspace sections; direct how-to questions can use relevant in-app Help where available. Organizer, Dashboard, conversation, memory, and media controls work together, but the options you see depend on your account and deployment.

Ask Friday to guide you

In an authorized FridayLocalAI conversation, try:

Give me a guided tour of FridayLocalAI focused on my first task. Explain Organizer, conversation, Dashboard, Agent, Model, and Help. Ask me what controls I can see and wait before moving to the next step.

Tell Friday what you can see and which step you have reached. If the instructions do not match your screen, say so; ask for the relevant Help topic or contact your operator. Friday cannot be assumed to see your screen or know your current settings.

Understand the workspace before adding complexity

Friday’s Workshop is the user-facing environment for working with FridayLocalAI. Its main regions have different jobs: Organizer helps you find work, the conversation holds the exchange, and Dashboard brings related context and controls into view. Layout controls let you show or hide regions, while Media Studio opens a dedicated area for configured media workflows. Your role and installation affect which controls are available. Learning those regions is more useful than memorizing a long list of features. Begin by locating the active conversation and confirming the intended project. Then identify the Agent and Model controls and find the information icon for a section you do not understand. If a panel is hidden, ask about its layout control rather than concluding that the underlying feature has been removed.

Projects give ongoing work a structure

A project provides a place to organize related activity. Folders can divide that activity into smaller areas, and conversations can focus on specific questions or deliverables. For a writing project, you might separate research, outlines, and drafts. For a technical evaluation, you might separate requirements, experiments, and review notes. These are suggested organizational patterns, not structures the application creates automatically. Use names that describe the work rather than the model used on a particular day. A conversation called “Compare document-review approaches” remains meaningful when you change models. Before attaching material or saving guidance, check that the intended project and conversation are selected. Good organization helps you find the right context; it does not replace the application’s access controls or authorize a different user to see the work.

Conversation is an iterative working process

A productive exchange usually begins with a goal, a small amount of relevant context, and a description of the desired result. Ask for an outline before a long draft if the structure is uncertain. Ask for assumptions before a recommendation if important facts are missing. After reading the response, give a concrete correction rather than simply requesting that it be “better.” The application can preserve the conversation, but saved text and verified facts are different things. A response may be fluent and still misunderstand the task. Compare important details with the original source and inspect any claimed result. When you return later, read the recent exchange and confirm what remains to be done. This makes continuity useful without treating every earlier statement as established truth.

Context has a purpose and a scope

Documents, prompt layers, and Memory Notes influence work in different ways. A document is evidence or reference material when its contents are available. A prompt describes how to approach a task. A Memory Note captures guidance or information within an applicable scope. Using the right mechanism helps avoid confusing a temporary request with a broader instruction. For example, an organization-wide writing preference may belong in a permitted broad scope, while the required format for one report may belong in its conversation. A scanned document might need a supported processing path before its contents can be used. Ask Friday to explain those distinctions, then verify the relevant screen or source. Do not infer a successful memory write or document review from a conversational assurance alone.

The model is one component of the platform

A local LLM supplies language capabilities, but an application also needs interfaces, storage, context handling, policy, and operational services. FridayLocalAI brings those pieces together so their roles can be understood separately. The requested model may not always be the active model for a response because Agent and capability routing can affect the choice. Response attribution is therefore important when comparing behavior. Tools are another separate layer. A model may explain a filesystem task without having permission to perform it. A Skill may describe a process without enabling the required operation. If you need an actual action, confirm that the capability is configured, permitted, and appropriate to the target. The evidence of completion is the result, not the confidence of the explanation.

A practical evaluation of fit

Choose one representative workflow and describe its starting material, expected output, and review criteria. For document work, that might be a summary with accurate source references. For writing, it might be a draft that follows a supplied outline. For creative work, it might be a completed artifact meeting a defined brief. Test the parts that matter to your use rather than assuming a general demonstration covers everything. Record where inputs are processed, which services are involved, and who operates the environment. A self-hosted AI deployment can offer useful control while also requiring maintenance and support. Bring those questions to an early-access conversation with the project. No public installer or purchase is required to begin that discussion, and no broad compatibility or performance claim should substitute for checking your intended workload.