> ## Documentation Index
> Fetch the complete documentation index at: https://docs.usefluency.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Fluency: Frequently Asked Questions About the Platform

> Common questions about Fluency — covering workflow discovery, AI-ranked opportunities, agent behavior, the Assistant, and data access.

This page answers the questions Fluency customers ask most often — covering how the platform works, what data it accesses, how agents behave, and how to get the most value from the platform.

## Getting Started

<AccordionGroup>
  <Accordion title="Do I need to set up integrations before using Fluency?">
    No. Fluency requires no integrations or engineering projects to start. You'll see first insights within hours of onboarding — no IT involvement or API connections required.
  </Accordion>

  <Accordion title="How long does initial workflow discovery take?">
    Initial discovery typically completes within a few hours. The platform continues updating continuously as new activity occurs, so your workflow maps become richer over time without any manual intervention.
  </Accordion>

  <Accordion title="What do I need to get started?">
    A Fluency account is all you need. Contact your implementation team or log in at [app.usefluency.com](https://app.usefluency.com). No IT projects, API connections, or data migrations are required before you begin.
  </Accordion>

  <Accordion title="Can I add team members to my Fluency workspace?">
    Yes. Go to **Settings > Users** to invite colleagues. You can assign different access levels depending on each person's role, so teams see the workflows and initiatives most relevant to them.
  </Accordion>
</AccordionGroup>

## Workflow Discovery & Work Explorer

<AccordionGroup>
  <Accordion title="What does Fluency actually analyze to discover workflows?">
    Fluency captures work activity signals across your organization to map process steps, handoffs, and decision points. It does not require direct database or API access to the systems involved — the platform builds process maps from observed activity patterns rather than direct system connections.
  </Accordion>

  <Accordion title="How current is the workflow data in Work Explorer?">
    Continuously updated. There is no manual refresh step — the platform keeps all workflow maps current as activity occurs. The timestamp on each workflow reflects the most recent activity detected.
  </Accordion>

  <Accordion title="Can I filter which workflows appear in Work Explorer?">
    Yes. Use the filter controls to scope your view by team, department, date range, or workflow type. Filters apply instantly and can be combined to focus on exactly the processes you want to analyze.
  </Accordion>

  <Accordion title="What is a bottleneck in Fluency's context?">
    A bottleneck is a step where the process consistently stalls — runs that pass through it take significantly longer or have a higher deviation or rework rate than average. Fluency highlights these automatically in the workflow detail view so you can investigate without manually comparing run data.
  </Accordion>
</AccordionGroup>

## Opportunities

<AccordionGroup>
  <Accordion title="How does Fluency calculate the confidence score for an opportunity?">
    The confidence score reflects how structurally consistent a workflow is and how well it matches known automation patterns. Scores above 85% indicate strong automation candidates with predictable, repeatable steps. Lower scores may reflect workflows that need more data to accumulate or process cleanup before automation is viable.
  </Accordion>

  <Accordion title="What is the difference between Agentic, Workflow, API, Process, and RPA automation types?">
    Each type reflects a different approach to automating a workflow:

    * **Agentic** — A fully autonomous AI agent handles the workflow end-to-end, making decisions and taking actions without human intervention on each run.
    * **Workflow** — A structured multi-step automation with defined triggers, branching logic, and handoff points.
    * **API** — A direct system-to-system integration that moves data or triggers actions between platforms.
    * **Process** — A redesign recommendation that improves the workflow without full automation — useful when human judgment is still required at key steps.
    * **RPA** — Robotic automation for UI-based repetitive tasks, typically used when no API or integration path exists.
  </Accordion>

  <Accordion title="How often are opportunities refreshed?">
    Continuously. As new workflow data arrives, opportunity scores and rankings update automatically. You don't need to trigger a recalculation — the Opportunities view always reflects the latest available data.
  </Accordion>

  <Accordion title="Can I dismiss an opportunity I don't want to pursue?">
    Yes. Use the dismiss action on any opportunity row. Dismissed opportunities are hidden from the default view but are not deleted — you can restore them from the dismissed view at any time if priorities change.
  </Accordion>
</AccordionGroup>

## Initiatives & Agents

<AccordionGroup>
  <Accordion title="What does an agent do when it encounters a case it can't handle?">
    The agent surfaces the run as an exception task in the Initiatives dashboard. You'll see the full context — the workflow run, the step where the agent stopped, and the agent's proposed action — so you can review it and choose to approve, override, or update the agent's handling rules.
  </Accordion>

  <Accordion title="Can I pause an initiative without losing run history?">
    Yes. Pausing an initiative stops new runs from being processed but preserves all historical run data, exception logs, and performance metrics. You can resume or reconfigure the initiative at any time without losing continuity.
  </Accordion>

  <Accordion title="How do I know if an agent is performing as expected?">
    The Initiatives dashboard shows cumulative run counts, exception rates, cycle time comparisons, and hours saved versus projections. You can also use the Assistant to generate a narrative performance summary for any initiative at any time.
  </Accordion>

  <Accordion title="What systems does an agent interact with?">
    Each initiative's detail view lists the required systems and the specific actions the agent takes in each one. Agents only interact with systems that were identified as part of the original workflow — they don't access systems outside the defined scope of the initiative.
  </Accordion>
</AccordionGroup>

## Assistant

<AccordionGroup>
  <Accordion title="What can I ask the Fluency Assistant?">
    The Assistant answers natural language questions about your workflow data — for example, "Which workflows have the highest cycle time this month?" or "Summarize the performance of the Invoice Processing initiative." It draws on the same data visible in Work Explorer, Opportunities, and Initiatives, so you can explore insights conversationally without navigating between views.
  </Accordion>

  <Accordion title="Does the Assistant take actions, or does it only answer questions?">
    The Assistant is a read-only, question-and-answer interface. It surfaces information and generates summaries, but it does not create, modify, or delete initiatives, opportunities, or workflow configurations. Any changes you want to make must be done through the relevant platform view.
  </Accordion>

  <Accordion title="How current is the data the Assistant uses?">
    The Assistant draws on the same continuously updated data as the rest of the platform. Answers reflect the most recent workflow activity and initiative metrics available — there is no separate refresh step required.
  </Accordion>

  <Accordion title="Why did the Assistant say it doesn't have enough data to answer?">
    The Assistant can only answer questions about workflows and initiatives that have been discovered and have sufficient activity signals. If a workflow is new or hasn't yet accumulated enough runs, the Assistant will indicate that there isn't enough data for a reliable answer. More activity over time will make its responses more complete.
  </Accordion>
</AccordionGroup>

## Data & Security

<AccordionGroup>
  <Accordion title="What data does Fluency access?">
    Fluency captures workflow activity signals — process steps, timing, handoffs, and team context — to build process maps. It does not store or process transactional business data, personal customer data, or financial records. See [trust.usefluency.com](https://trust.usefluency.com) for full details on what Fluency accesses and how it's handled.
  </Accordion>

  <Accordion title="Is Fluency compliant with industry regulations?">
    Fluency's compliance documentation — including security certifications and data handling policies — is available at [trust.usefluency.com](https://trust.usefluency.com). If you operate in a regulated industry, your implementation team can walk you through the relevant documentation for your compliance review.
  </Accordion>

  <Accordion title="Where is my data stored?">
    Fluency uses enterprise-grade cloud infrastructure. For specific data residency requirements — such as data stored in a particular geographic region — contact your implementation team.
  </Accordion>
</AccordionGroup>