#ai

15 questions

Does Corcava have an AI assistant?

Yes! Corcava includes an AI-powered business assistant that answers questions about your data in plain English. Ask questions like 'How many hours did my team work last week?' or 'Which project was most profitable this month?' and get instant answers. The AI understands natural language, recognizes team member names, and works with your time tracking, financial, and project data. Available to organization managers and administrators to protect sensitive business information.

Is my data safe when using the AI assistant?

Yes, your data is secure. The AI assistant processes queries in real-time using your current data and doesn't store your business information. All queries respect your team permissions and data isolation. While OpenAI processes the queries to provide intelligent responses, your business data is not retained by OpenAI. Conversations are private to each user and not shared with other team members.

What can I ask the AI Business Assistant?

You can ask the AI assistant about time tracking (hours worked by team or project), project profitability (revenue, costs, profit margins), client billing (unbilled time, total revenue), and financial comparisons. Use natural language like 'Show me time tracked for Project X last week' or 'Which team member was most active yesterday?' The AI understands date expressions like 'last week' and 'this month' and recognizes team member names even if you use first names or nicknames.

Can AI analyze my time tracking screenshots?

Yes! Corcava's AI Screenshot Analysis lets you ask any question about your screenshots in natural language. For example, ask 'Was this person watching YouTube?' or 'What applications were being used?' and get instant AI-powered answers. The AI can identify applications, detect activity patterns, classify work vs. entertainment, and provide summaries. Access it from the Screenshots page by clicking 'Analyze Screenshots'.

What questions can I ask about my screenshots?

You can ask virtually any question about what's shown in your screenshots. Popular examples include: 'Was this person watching YouTube or other video sites?', 'Find screenshots showing social media usage', 'What applications were being used?', 'Is this productive work or distraction?', 'Flag screenshots that don't look like work', and 'Summarize the work activity'. The AI understands natural language and provides answers with confidence scores.

How does the token system work for screenshot analysis?

AI Screenshot Analysis uses tokens from your account balance. The rate is 1 token = 100 screenshots analyzed. Before starting an analysis, you'll see how many screenshots will be processed and how many tokens are required. Purchase tokens from the Billing page. The AI uses smart clustering to group similar screenshots, reducing the number of API calls needed and optimizing your token usage.

Who can see my screenshot analysis results?

You control who sees your analysis results through visibility settings. Choose 'Personal' to keep results visible only to yourself, or 'Team' to allow managers and admins to view them. You can only analyze screenshots you have permission to view - your own screenshots, or team member screenshots if you're a manager or admin. All analysis activities are logged for security.

Can I connect AI assistants like Claude or Cursor to Corcava?

Yes! Corcava supports the Model Context Protocol (MCP), allowing you to connect AI assistants like Claude Desktop, Cursor, Windsurf, and Continue directly to your workspace. Your AI can read task details, create tasks, track time, add comments, and more. To set it up, go to Settings > Integrations > Public API to generate an API key, then configure your AI assistant with the MCP server URL (https://app.corcava.com/mcp) and your API key.

How does AI task estimation work?

AI task estimation uses your linked GitHub or GitLab repository to suggest realistic development time (or an implementation approach) for a task. Connect a repo in Settings > Integrations (same as for the Coding Report), then from a task on a project board use the Analyze or Estimate action. The AI examines your codebase and returns estimated hours, complexity, and a short explanation. You can apply the estimate to the task so it shows on the board.